HORACE W. CARFENTIER
LECTURES
UPON
THE ASSYRIAN LANGUAGE,
AND
SYLLABARY.
ft
LECTURES
UPON THK
ASSYRIAN LANGUAGE,
AND
S Y LL A B A RY;
DELIVERED TO
THE STUDENTS OF THE ARCHAIC CLASSES.
BY
Rev. A. H. SAYCE, M.A.,
Deputy Professor of Comparative Philology, Oxford,
T
Multse terricolis ling-uae, coelestibus una,
LONDON:
SAMUEL BAGSTER AND SONS,
15, PATERNOSTER ROW.
\All rights reserved.'] »
1877.
/^-^^^xx^fe^
^77
CONTENTS.
PAGE
Preface ... ... ... ... ... ... ... vii
Lecture I. Introduction ... ... ... ... ... i
Lecture IL The Syllabary ... ... ... ... ... 9
Lecture III. The Syllabary continued ... ... ... 23
Lecture IV, The Propagation of the Syllabary ... ... ... 37
Lecture V. Assyrian Phonology ... ... ... ... 45
Lecture VI. The Pronouns ... ... ... ... ... 63
Lecture VII. The Verb ... ... ... ... ... 77
Paradigms of the Assyrian Verb ... ... ... 106
Lecture VIII. Assyrian Syntax ... ... ... ... 114
Lecture IX. Affinities of Assyrian and the Origin of Semitic Culture ... 134
GiSlS4
PREFACE.
HE following Lectures form part of an experiment which took practical
shape through the unwearied exertions of Mr. W. R. Cooper, the
Secretary of the Society of Biblical Archaeology. Classes in Egyptian and
Assyrian were started in the rooms of the Society in the spring of 1875 ;
Mr. Le Page Renouf superintending the first, and myself the second. The
three first Lectures on the syllabary embody the substance of the Lectures
delivered in 1875, before an audience which averaged some thirty students ;
the remaining Lectures occupied the spring of 1876, the second year of the
experiment. The success which attended it leads to the hope that English
schools of Egyptology and Assyriology may be permanently formed, and the
study of the monumental languages of the great nations of antiquity placed
on the same footing as the study of Hebrew.
For Assyrian, two classes of students are urgently required. One, whose
eyesight and practice shall enable them to copy the minute characters of
the Assyrian tablets with photographic accuracy ; the other, who shall bring
to the task of decipherment all the varied stores of Semitic philology and
learning. Of course both classes must, to a certain extent, intermingle their
acquirements ; the philologist ought to be able to control the epigraphist,
viii Preface.
and the epigraphist to have some knowledge of Semitic philology. But in
these days of divided labour, and in a subject of so vast extent as Assyrian
decipherment, it is not necessary, indeed it is rarely possible, that the two
specialties should be united in the same person. Ordinarily, the philologist
must content himself merely with that knowledge of epigraphy needful for
his purpose, the epigraphist with that knowledge of philology needful to
guide him in his readings. Inasmuch, however, as the study of Assyrian is
a monumental one, the philologist will have to be an epigraphist to a far
greater extent than is the case with the classical scholar.
There is no doubt a good deal in the following Lectures which may have
to be corrected by subsequent discovery. Such must always be the case with
a progressive study. Nevertheless, the main outlines of Assyrian grammar
have now been sketched with clearness and certainty, its main problems have
been solved, and the details alone left to be filled in. I, for one, believe that
the day is not far distant when it will be recognised that a knowledge of
Assyrian is as important for comparative Semitic philology as is a knowledge
of Sanskrit for the comparative study of the Aryan languages.
A. H. SAYCE.
Queen's College, Oxford, yune 6t/i, 1877.
PHILOLOGICAL LECTURES
ON THE
ASSYRIAN LANGUAGE.
LECTURE L
On the Study of the Assyriau Language.
T is with mingled feelings of gratification and diffidence that I come
before you this evening to open a series of lectures, the character and
object of which are new and even revolutionary in the history of our studies
and education. For the first time in this country an attempt will be made
to found a system of instruction in languages, which it has been the glory of
the present century to recover from the past, which are clothed with all the
modern interest that attaches to the great problems of the development of
civilization, and which demand, not mere memory or dependence upon the
authority of others, but the new methods of patient scientific induction.
Thanks to the exertions of the indefatigable Secretary of the Society of Biblical
ArchcBology, Mr. W. R. Cooper, my colleague Mr. Le Page Renouf and myself are
enabled to bring before your notice classics more ancient than those of Greece
or Rome, or even Judea — classics, too, which are written on contemporaneous
monuments, and must be spelled out, as it were, from the lips of a living
people — explaining the details of their grammar and idioms, and the key
2 LECTURE I.
which has unlocked their secrets. The knowledge of all this has hitherto been
confined, like the sacred learning of Egyptian priests, to a small band of
workers, from whom the world has been content to accept the startling results
which have from time to time awakened its incredulity or excited its interest ;
and no endeavour has yet been made in England to bring the languages and
the literature of the pioneers of civilization out of the mysterious shadow-
land of the specialist into the commonplace light of the lecture-room and
the school. Shall I be considered presumptuous if I say that the courses of
lectures which I have been permitted to inaugurate this evening mark an
era in national education ? I cannot express the gratification I feel at the
attendance which I see before me, so large beyond my boldest expectations,
and so encouraging to the success of our work. A few years back the languages
and the literature, which will be the subject of our studies, lay forgotten and
unknown under the rubbish of centuries, or in the dusty corners of European
museums ; still fewer years ago they were but a sealed book to all but one or
two daring scholars who alone were attempting to penetrate their contents.
Already they stand on a level with the manifold subjects of human know-
ledge which are taught and learned, and the students who have gathered this
evening to help us in founding schools and educational courses of Assyrian
and Egyptian philology, are a token that a fresh start has been made in the
education of the country, and a fresh realm of conquest opened out before the
mind.
For, we must remember, the study of Assyrian and Egyptian philolog}^
differs in several very essential points from the studies with which we are
usually familiar ; and since the method by which it must be learnt is a new
one, a new method also must be devised for teaching it. Firstly, and
especially, the teacher and the pupil must both alike be learners, and the
difference between them is one of degree only, and not of kind. The teacher
is but a little in advance of the pupil", but feeling a way, as it were, for the
latter, and even in the act of teaching, is making fresh discoveries, and
rectifying old conclusions. There is no authoritative standard to be referred
to, no tradition to be appealed to, no dictionary to be consulted ; all must
be worked out by the laborious comparison of texts, by extensive knowledge
of cognate languages, by ready combination and hypothesis, and by the
trained judgment of scientific research. In short, the decipherer is as much
ON THE STUDY OF THE ASSYRIAN LANGUAGE. 3
a discoverer as the man of science, the chief distinction between them being,
that whereas the man of science has now a tradition, an authority, a standard
to look up to, the decipherer is still engaged in creating one, and elaborating
out of his own experience a method for others to follow. Such a pursuit is
thoroug*hly in harmony with the independent and inquiring spirit of our own
age — indeed, we can hardly imagine it arising at any previous period; and the
work we have before us is none other than to cast into what we may call
an educational mould this embodiment of our nineteenth century spirit. It
is to do for language and literature, for littered humaniores, in fact, what was
done two or three centuries ago for science. Such an attempt has perhaps
never been made before, unless we go back to the time when Athenian
sophists and orators were struggling to find out the force and meaning of
the words they uttered and framing a Greek Grammar. Since then the
literary and linguistic education of Europe has been confined within the
limits of a traditional system. The Romans made Greek the basis of in-
struction, and so built up a grammar and literature of their own, which have
formed the groundwork and staple of the education of later times. There
have always been a framework and method to fall back upon, accidentally in
existence if you like, but still in existence it was ; and the young mind was
accordingly kept in the leading-strings of the past, and taught to lean upon
a cramping authority. To feel and exert its own powers, to educate itself in
the truest and fullest sense of the word, is a task that has been reserved for
our own days. In the decipherment of the ancient classics of Babylon and
Egypt, in the gradual recovery of that Oriental past, which is so all-important
for the history of intellectual development, as much as in discoveries of
science, the servant is not above his master ; and the reason is the same in
each case, for the method which we have to employ is no less the comparative
method of inductive science than that of the chemist or geologist.
The second point in which the subject of our lectures differs widely from the
subjects of the ordinary curriculum, is its contemporaneous character. We
have not to deal with the late MS. copies of illiterate or careless scribes, but
with the very documents which came from their authors' hands. It is true that
many of these are copies or editions of older records, so that the purity of the
text may still exercise the intelligence and call forth the reasoning powers of
the scholar ; but, nevertheless, they were written when the language was yet
2*
LECTURE I.
living and spoken, and their very faults are a valuable evidence of the state of
the language and its speakers at the time they were inscribed. Epigraphy is
one of the studies which has grown up of late years, and from the nature of
things it must always take but a subordinate place in the study of Latin and
Greek; but epigraphy, in the sense of the study of contemporaneous ^records,
is the sum and substance of our Assyrian and Egyptian researches, which are
essentially occupied with the decipherment of contemporaneous inscriptions.
This contemporaneousness is of inestimable importance, even from an educa-
tional point of view. To find oneself face to face with the writers we study, to
reach them through no later channels, more or less fallacious, but to speak to
them as to living men, removes that artificial unreality which as those who
have had anything to .do with education know too well exercises so fatal and
dulling an influence upon the mind. From other points of view besides the
purely educational one, the manifold advantages that result from having to
deal with contemporaneous documents need not be dwelt upon. One only will
I single out, as that has a special bearing upon the immediate subject of these
lectures. I mean the opportunity thus afforded us of tracing the growth and
history of the language from the period when it first becomes literary, down to
its closing epoch. It is only in this way that we can ever really come to
know a language, and the certainty with which we can do so in Assyrian makes
the latter invaluable not only for Semitic philology in particular, but for com-
parative philology in general. Already, as will be noted in the course of these
lectures, light has been thrown by Assyrian upon some of the obscurest points
of Semitic grammar, while the discovery of Accadian, the oldest form of agglu-
tinative speech, is likely to create a revolution in Turanian studies, and to
solve not a few of the problems of the science of language. Elucidation of
Semitic philology necessarily brings with it elucidation of the Old Testament
writings ; and questions like that of the possibility of a Hebrew construction,
or the probability of a corrupt reading, can only be decided by monuments
inscribed in a kindred dialect at a time when Hebrew was still a spoken
tongue.
The third and last point to which I shall advert wherein these lectures
and classes are introducing a new educational force, is their testimony that
there is something worth learning besides the time-honoured subjects of school
and University training. We are apt to become narrow and conventional in
ON THE STUDY OF THE ASSYRIAN LANGUAGE. 5
our habits of thought, and to regard everything with which we are unfamiHar
as barbarian. It is the old error of the Greek and Roman over again. We
must learn that there was a culture and civilization five or six thousand years
ago on the banks of the Nile and Euphrates which would compare favourably
with that of our forefathers but three or four centuries since, and that the
literary productions of these ancient people are as admirable in their own way
as the masterpieces which have stereotyped our canons of taste. The merits
of this or that study, of this or that method of education, are but relative ; and
it may yet turn out that a study and method which require the free and
unchecked exercise of our mental powers, which demand all the qualities
on which the man of science prides himself, and which call us back to
originals rather than to copies, are more in harmony with the needs of a
future generation than the studies and the methods which now possess our
minds.
There is one fact, however, which must not be blinked, and it is a fact that
meets us on the very threshold of our researches. The languages we propose
to study are concealed and buried beneath a pyramid of strange and uncouth
characters. Before proceeding a single step, we have to load our memories
with an endless and intricate syllabary. The preliminary toil is very great,
and it is well that this should be realised at the outset. But let us remember
that nothing good and sound has ever been achieved without trouble, and that
if we mean serious work we cannot expect to find everything smooth and easy-
going. The life of the scholar and the life of the dilettante are two very
different things ; but the dilettante never accomplishes anything except the
selfish art of killing time. Do not, then, be frightened by the multitude of
polyphonous characters which have to be learned before we can interpret the
Assyrian inscriptions to any purpose, or the long lists of hieroglyphics which
Mr. Le Page Renouf will require you to commit to memory. These diffi-
culties have been overcome by others before you, and a time will come when
the acquisition of a new character will bring with it a real pleasure. But let
us not be deceived into thinking that we can study Assyrian and Egyptian
without first mastering the characters in which these languages are written.
Transliteration may be a good help, but it will be a broken reed to lean upon
alone. Confining myself to Assyrian I must recall the fact that the existence
of polyphones necessitates a combination of the decipherer and philologist.
LECTURE I.
We cannot speculate on the meaning and affinities of a word unless we know
how to read it, and we cannot know how to read it unless we also know what
value to select in any given case out of the many possible ones a character
may bear. All that I can do is to lighten the burden of learning this ponderous
syllabary by explaining its origin, and setting forth the rules to be followed in
reading the inscriptions ; and this I shall try to do in the first two or three
lectures. But I cannot prevent the task from being a distasteful and irksome
one, and from having perforce to be gone through.
There is yet another point on which I would remove all chances of a false
impression. Just as the preliminary labour of learning the syllabary must be
no holiday amusement, so also must the study of the Assyrian grammar
be thoroughgoing and scholarly. We must have no slovenly and merely
approximate translations ; and while in the course of these lectures I shall
keep your attention fixed upon the principal outlines and main facts of the
Assyrian grammar, I shall at the same time insist upon those small niceties
and distinctions which are apt to be overlooked by the hasty and superficial
student, but which stamp and distinguish a language more than anything else,
and prevent the translator from losing the idiom, and with that the sense and
meaning of the original. It will often be found that the signification of
important passages depends upon this accuracy of scholarship. In a Semitic
language it is the verb in which these niceties are liable to be ridden roughshod
over, to the detriment not only of the study of the language itself, but even
more of the force and drift of the text. The conception which underlies the
Semitic verb is so radically different from that to which we are accustomed in
our own family- of speech, that careful investigation alone can really discover
its various forms and uses. It is only within the present century that any true
knowledge of the Hebrew verb has been arrived at, and passages of the Old
Testament, which before seemed hopelessly obscure, cleared up and assigned
their true meaning. By way of illustration, take, for instance, the first few
verses of Genesis. We all know well the way in which they are translated in
our authorised version. But when we give each of the tenses employed in them
the peculiar force and signification which modern research has shown them to
have, ascribing to the perfect ^^"^I1 its sense of completion, to the perfect,
divided from the copula by the subject, its pluperfect value, to the participle
its meaning of continuance, to the imperfect with waw consecutive its sense of
ON THE STUDY OF THE ASSYRIAN LANGUAGE, 7
subordination and incipiency, what a change is made in the meaning of the
whole passage, what fresh vividness is given to the picture ! "In the beginning
God hewed out the heaven and the earth : now the earth had been waste and
desolate, and darkness on the face of the deep ; and the Spirit of God was ever
brooding on the face of the waters ; and God said," etc. Here a new significa-
tion has been put into the verses by modern research — a new life breathed
into them by a more accurate knowledge of the Hebrew verb. Now, just as
we cannot afford to study Hebrew without thoroughly acquainting ourselves
with its use of the tenses, so also ought it to be in Assyrian. Merely to be
able to give a sort of rough guess at the signification of a sentence, setting
down what we believe to be the substance of it, and overlooking all the finer
points of grammatical idiom, is not to be a translator in the proper sense of the
word. Before everything else, grammatical accuracy is absolutely requisite.
When once we are sure of the grammar of a passage, the lexical difficulties
will soon disappear. Of course this close attention to what has been con-
temptuously termed "the minutiae of Assyrian grammar," is not likely to be
popular. But again, let us remind ourselves that we are here not to be
dilettanti, but scholars. Following the late Dr. Hincks, I have stood almost
alone in the endeavour to trace the meanings and usages of different forms of
the Assyrian verb, protesting against the rough-and-ready process that would
lump them all together indiscriminately, and place Assyrian grammar on the
same footing as was Hebrew grammar before the investigations of modern
scholars. My Assyrian Grammar was an attempt, however inadequate and
humble, to do for Assyrian what Ewald or Olshausen have done for Hebrew ;
and I am glad to think that the attempt has not been altogether a failure.
One by one my colleagues in the study of Assyrian have adopted my views, at
all events in principle ; and my friend, Dr. Schrader, who once expressed his
unqualified and emphatic dissent from them, has so far come to agree with me
that he acknowledges the difi'erence between us to be now one of name only.
The system of grammar propounded in the present lectures will be in accor-
dance with the principles laid down in my " Grammar." Much that I have
said there was polemical, and now therefore superfluous, while, in some other
respects, it has had to be supplemented or modified ; it will, therefore, be the
substance of my maturer conclusions, based upon a fuller investigation of the
inscriptions, which will be given in the lectures it is my privilege to deliver
8
LECTURE I.
before you. You will, I think, be convinced that the subtle and extensive
machinery of the Assyrian verb, so far from being one to make " Semitic
philologists shake the head," is in full harmony with that of Hebrew or
Ethiopic or Arabic, and the surest token which we possess of the pure
Semitism of the Assyrian tongue. The perfection is only what we should
expect from a language so complete and primitive in character and age, and
throws a flood of light upon a long misunderstood part of Semitic philology.
LECTURE II.
The Syllabary.
|T the very threshold of his Assyrian studies the beginner is met by the
most repulsive, but, nevertheless, an indispensable part of what he
has to learn. Nothing is harder than to familiarise oneself with a new
character which has not been learnt in childhood, and so become, as it were,
part of the furniture of the mind. One's own language looks strange and
difficult when written in a foreign alphabet, and even a page of the " Fonetic
Nuz" requires some spelling out. But the repellent difficulties encountered
in acquiring the knowledge of a new character are increased a thousandfold
when the number of distinct characters amounts to four or five hundred, each
of them possessing more than one value. Yet such is the case with the
Assyrian syllabary ; and it is well to state the full difficulties of it at once.
The difficulties, however, will not prove insurmountable ; and the best proof of
the possibility of getting over them is that the powers of the numerous signs
of the syllabary have all been made out one after the other by patient
decipherers during the last twenty years, and that fresh scholars have from
time to time been entering the field, undaunted by the task of mastering the
Assyrian mode of writing. A Spanish author called his Basque Grammar
" The Impossible Vanquished," and here we have to do with another
" impossible " which can be vanquished with equal facility.
Now I cannot, of course, impress the Assyrian syllabary upon your
memories, and make each character as familiar to you as the letters of our
10 LECTURE II.
own alphabet without further trouble ; but I can explain the way in which
this cumbrous system of writing originated and grew up, and so (as I hope)
can lighten your labours in learning it. The reason of a thing is more than
half the whole ; and when we are arrived at years of discretion, and have
ceased to repeat our lessons by rote, like a cage of parrots, our memory is
enormously aided by being accompanied by the intelligence. To understand
is to remember.
What I shall try to do, therefore, in this second lecture, is to point out the
basis and system upon which the Assyrian syllabary rests, or, in other words,
to give a sketch of its origin and development. Before doing this, however,
I would impress upon you the necessity of learning some portion at least of
the syllabary before attempting to read the inscriptions. To depend upon
the transliteration of another is always unsatisfactory ; doubly so in the case
of a syllabary, the characters of which have more than one value. If nothing
more, at all events a knowledge sufficient to control the reading of an
inscription is requisite. For merely comparative purposes, indeed, we may
take on trust the transliterated examples of Assyrian grammatical forms ; but
if we want to translate the inscriptions for ourselves, and to enter into the
niceties of the language, we must be able to read them as we would a work in
Hebrew or Greek. It is not necessary to know all the characters of the
syllabary and the manifold values of each ; indeed, many of them are still
unknown, and others occur only in a single passage, or in the bilingual
tablets. But it is necessary to be acquainted with those in most frequent
use, and the best preliminary to the study of Assyrian would be to learn as
perfectly as possible the different characters and ideographs, with all their
varying powers, which are prefixed to the first volume of Mr. Norris's
dictionary. By going over them frequently, the eye w^ould soon become
habituated to their forms ; and if, after gaining a first general knowledge of
them the student would take some texts and endeavour to write these down in
English letters, always keeping the syllabary at his side for reference, he
would quickly find himself making astonishingly rapid advances. After a
time, a complete syllabary like that given in my Elementary Assyrian Grammar
and Reading Book (Bagster and Sons) should be used, and fresh signs and fresh
values would thus be continually imprinting themselves upon the memory. Up
to the last, however, he will discover that he cannot altogether dispense with
THE SYLLABARY. 1 1
such an aid, supplemented by himself, as it certainly will be, in the course of
his own researches. It is happily not needful to burden the already over-
weighted memory with the load of the whole Assyrian syllabary ; characters
of rare occurrence can be hunted up whenever they are met with, and safely
left to the keeping of a written memorandum. Even the Assyrian scribe of
Assur-bani-pal's day, who had nothing else to occupy his thoughts, did not
profess to recollect all the signs of his own system of writing. He sometimes,
in copying, mistook characters of similar form for one another, and often came
across a character stamped in the old style, of which he did not know the
later equivalent ; while the best explanation that can be afforded of the fact
that the syllabaria often give but a few out of the many values possessed by a
particular character is that the writer could not remember at the time any
more than those he has set down.' When there is so much that is important
to remember, it is unwise to load our memories with what is needless. I do
not think, however, that it is advisable to depend long upon texts already
transliterated. No doubt it is useful at first to have an Assyrian character
before you with the transliteration of it underneath ; but crutches of this kind
should be thrown away as soon as possible. The requisite familiarity with
the syllabary can never be acquired, so long as it is instinctively felt that help
is close at hand without any trouble of thinking or searching for oneself ; we can-
not be sure that we really know the power of a character if the eye can take in
both the character and its power at a single glance. There will be a tendency,
too, to take for granted the particular reading of the particular translator whose
transliteration we are using, and prepossessions of this sort are among the
hardest things to eradicate. A certain character ( **^| f>p ) has, in Assyrian,
the two common values of saq and m, and when we meet with the word
t^^ '^||>fp >-Bt| " oracle " we are likely to assume, without further question,
that it must be read pi-ris-tii (from :i^")D) if Oppert has been the guide we
have followed, and pi-sak-tn (from ^_^) if Smith and Schrader have been
the first to initiate us into the mysteries of Assyrian. The Assyrian student
ought, above all things, to be independent, and it is only by mutual criticism
that the interpretation of the inscriptions can progress. Transliterations of
' This of course does not exclude another fact which seems to account for the selected number of values
assigned to a character in the syllabaria, namely that the scribe in compiling them went through some
Accadian text, setting down those values, and those values only, which a particular character bore in that text.
12 LECTURE II.
texts arc chiefly useful after an acquaintance with the syllabary has been
assured. It saves time and trouble to be able to read off an inscription in a
character less complicated than the Assyrian, and if we are thoroughly
acquainted with the native characters, an error or uncertainty in the reading
can be detected at once.
After these practical hints, we may now go on to the theoretical part of
the subject. And the first question that starts up, strange as it may appear,
is, what is the Assyrian syllabary which has to be learnt ? Both the
inscriptions and the printed texts offer us different types of writing, which
seem to differ wholly from one another. We may be perfectly familiar with
the inscriptions of Assur-bani-pal and yet utterly unable to make out those of
Nebuchadrezzar ; we may be able to read those legends of Sargon's which
are lithographed in Western Asiatic Inscriptions, Vol. I., pi. 36, or Layard,
pi. ;i,;^, and yet find ourselves hopelessly puzzled over the same monarch's
inscription on the Cyprian monolith (W. A. /., Vol. III., 11), or the obelisk
of Samas-Rimmon ; and a knowledge of all these may leave us uncertain of
the values of the characters on the contract-stones of Babylonia, or the clay
bricks of ancient Chaldea. Assyrian writing presents itself to us in at least
four different forms, and sooner or later we shall have to acquire a certain
amount of knowledge of all these four. First and foremost comes w4:iat we
may call the Archaic form of writing, out of which the others have been
gradually developed and simplified. The oldest inscriptions of which we
know are written in the Archaic type, and they all come from the primitive
cities of Chaldea. The numerous literary works copied and translated into
Assyrian by order of Assur-bani-pal or his predecessors were originally
inscribed in this style of cuneiform ; and it occasionaily happens that the
original character is reproduced in the copy through the scribe's ignorance of
its later Assyrian representative. A further development and simplification
of this Archaic cuneiform is generally termed the Hieratic. The contract
stones to which I alluded above may be said to be written in it ; so also is the
Cyprian inscription of Sargon and the obelisk records of Samas-Rimmon.
Though a modification of the Archaic, it preserved the old forms of the
characters much more closely than did the Assyrian, and was therefore used
as a sort of black letter for ornamental purposes at the court of Nineveh. It
is this use which has induced decipherers to call it Hieratic. Not very
THE SYLLABARY. 1 3
dissimilar to the Hieratic was the Babylonian cuneiform of Nebuchadrezzar
and his cotemporaries ; and the syllabary employed for the Assyrian
transcripts of the Persian inscriptions at Behistun and elsewhere, as well as
the selected syllabary of the Amardian or " Medo-Scythic " texts, is but a
simplified edition of the Babylonian.' Distinct from all these, and simpler
than any of them, except that of the Persian period, is the Assyrian, properly
so called, which is found on the great mass of Assyrian monuments, from the
sixteenth or fifteenth centuries b.c, down to the fall of the monarchy in the
seventh. It is this kind of cuneiform which has been taken as the type and
pattern of all the rest, and published in printed books. By far the largest
part of the inscriptions we possess are written in it ; and when we speak of
learning the Assyrian syllabary, accordingly, we mean that syllabary which
was specially and generally used in Assyria itself. After we have thoroughly
learnt this syllabary, the other four styles of writing may be acquired without
much additional labour : when once the Assyrian characters and their powers
have been imprinted on the memory, a little care and patience will show the
student how closely they are related to the corresponding Archaic, Hieratic,
and Babylonian; how, indeed, their descent from these may in a certain sense
be traced.
The cuneiform characters are degenerated hieroglyphics, like the Chinese
symbols or the Demotic writing of ancient Egypt. A fragment of a tablet
in the British Museum gives some of the primitive hieroglyphics side by side
with the cuneiform characters which have been corrupted from them. Thus,
the representation of a "comb" '^^~2^^ or S|\I ^T is given as the original of
^[^^|, Assyrian ^^j. Difficult as it generally is to discover any likeness
to a visible object in the signs of the Assyrian syllabary, we have often
only to trace them back to their Archaic originals to see how a particular
character came to stand for some particular object or idea. Primarily,
therefore, every character denoted some object or conception ; and we can
thus understand how it came about that the characters of the Assyrian
syllabary might be used as independent ideographs or hieroglyphics, as well
' In my Paper on the Languages of the Cu7ieiform Inscriptions of Elain and Media, in the Transactions
of the Society of Biblical Archceology, Vol. III., 465-485, will be found reasons for the use of the term
"Amardian." As " Protomedic," however, has now met with general acceptance, and is substantially
correct, it will be employed throughout the rest of these Lectures.
14 LECTURE II.
as mere unmeaning syllabic sounds. If we could come across any specimens
of the earliest attempts at writing in the Euphrates valley, we should expect
to find them consisting altogether (or, at all events, for the most part) of
ideographs ; and, as a matter of fact, the brick legends of the Chaldean
kings, as well as the old astrological tablets, contain many more ideographs
than meaningless phonetic syllables. In the Assyrian period, on the other
hand, the ideographic use of the characters was not common, except in
special cases, and even here, as we shall see, phonetic complements were
ordinaril}^ added to them to show how they were to be read. In fact, such an
ideographic use of the characters came to be mainly due to the desire of
abbreviation, just as we write viz. for "namely," + for " plus," i.e. for "that
is," and so forth. The clay tablets which served in the place of books were
necessarily of small size, and this want of space will explain the continued
employment of ideographs as well as the minuteness of the writing upon them.
In the larger inscriptions on stone, the use of ideographs may be accounted
for by the wish to end the line with the end of a word ; and in this way
ideographs in Assyrian played the part of littercB dilatabiles in Hebrew. How
long the primitive hieroglyphics took in passing into the cuneiform characters
it is impossible to say. We shall see that there is good reason to suppose
that writing on papyrus preceded writing on clay in Babylonia as well as in
Egypt ; and, while papyrus could be used for this purpose, there was nothing
to prevent the original picture-writing from being preserved without cor-
ruption. With the introduction of the "burnt bricks" or laterctdcB codiles, as
Pliny calls them, the case was altered. Angles had to supersede curves,
circles to make way for straight lines. Picture-writing on any extended scale
ceased to be possible ; and the impress of the style upon the wet clay caused
each line to assume a wedge-like form, the broad triangular base terminating
in a thin point. We still possess specimens of writing in which the transition
from the hieroglyphic to the cuneiform period is taking place. The hiero-
glyphics have ceased to be pictures, even more so than is the case with
the Egyptian Hieratic, but the characteristic wedge has not yet appeared ;
the lines are still drawn of the same breadth throughout ; they are still
joined one to the other, and are still able to be circular. When once,
however, the wedge-shaped characters, losing more and more of their original
hieroglyphic form, had come into vogue, the superior quickness and ease with
THE SYLLABARY. I 5
which they could be written soon made them universally prevalent. The
literary activity, of which Chaldea was now the centre, made rapid writing of
great importance, and so this cursive hand, as we may term the cuneiform,
came to be exclusively used. Further simplication was of course only a
matter of time.
Now a hieroglyphic system of writing has to represent ideas as well as
objects, and just as language expresses the spiritual through the veil of
material metaphor, so hieroglyphic writing must symbolise ideas by means of
objects. Thus two legs may denote "walking," a hand nvAy denote "to seize,"
breath may denote "the soul." But it is plain that the same object may
represent more than one conception; two legs may stand for "going,"
"running," "standing," "support," and even "growth," as well as for
"walking." Every hieroglyphic, therefore, may be pronounced in a variety
of different ways, according to its significations; and since few languages are
so poor as to be without synonymes, it may be pronounced in more than one
way, even when the same thing is meant. This will certainly be the case
where the same hieroglyphic system of writing is used by tribes who speak
different dialects ; the Chinese, for instance, have one set of written charac-
ters, but the variety of idioms spoken in the empire cause the same character
to be sounded in one way at Canton, and in another at Pekin. So long
as the writing continues to be purely hieroglyphic, all this produces no
confusion ; on the contrary it facilitates intercourse and civilisation. But
wherever there is any pretence to progressive culture, no writing can long
continue purely hieroglyphic ; although the native proper names may all be
significant, and so reducible to hieroglyphic representation, it is quite other-
wise with foreign proper names, and, however much a people may wish to
confine themselves within their own boundaries, like the Egyptians of the Old
Empire, or the Chinese of to-day, they will find themselves brought into
contact with unallied nations, and compelled to chronicle their foreign as
well as their home policy.' Sooner or later these troublesome proper names
will have to be written phonetically, that is, in hieroglyphics which are void of
meaning, and so have ceased to be hieroglyphics, but are sounded in some
' As in Egypt, the oldest names given by the early Accadian inhabitants of Babylonia to their neighbours
were of native origin like numma, Elam, " high ;" ^ubarti, Syria, of the same signification, and not the names
by which those neighbours called themselves.
l6 LECTURE II.
particular way. Out of the different pronunciations which can be attached
to a certain character, it will be necessary to select some which it shall
represent when used simply as the symbol of a particular phonetic power.
The obvious course, under these circumstances, might have seemed to select
a single phonetic power and attach it invariably to a character when non-
significant ; but obvious as such a procedure appears to us, it was not so
obvious to the old inventors of writing, and in numberless instances they
allowed a character to carry more than one value. What value, however,
was meant in any particular case was pointed out in several ways. In
certain combinations, a polyphone had always to be read with one special
value and no other, or the pronunciation attached to some object or idea
which seemed suitable to the individual or country denoted was chosen, or
again, the pronunciation was determined by the vowel of the character pre-
ceding or following, supposing that the syllable was an open one. Thus we
know that the second character in the name of Gtmgunuv must be sounded un,
because the vowel of the first syllable is u; and we might have concluded,
without the additional help of a gloss, that the second character ( i->^| ) in
the name of Ansan or south-western Elam was to be pronounced sa, and not
du, on account of the last syllable beginning with a.'
Of course the application of the merely phonetic employment of the old
hieroglyphics would be extended as soon as its convenience had been found
out. One of the first uses to which it would be put would be to express the
pronouns, which must have been a sore puzzle as long as there was only a
picture-writing to draw upon. In the Semitic and Aryan families, indeed,
the whole grammatical machinery of the language, the nerves and blood-
vessels that give life to the bare skeleton, would have been an equal puzzle.
Even the prepositions would have wanted some other mode of representation
than that of hieroglyphic writing. But the inventors of the cuneiform were
neither Semites nor Aryans, nor did they speak an inflectional language of
any kind. I should not, indeed, like to go so far as to say that the invention
of a purely hieroglyphical system of writing is inconceivable among those who
speak what are called inflectional languages, more especially as the Hamath
legends seem to show that an independent hieroglyphic system of home
' Since the character in question had the signification of "going" when used as an ideograph, it would
seem that sa meant " to go" in the language of Susa.
THE SYLLABARY.
17
growth was in use among the inhabitants of northern Syria, while the
grammar at least of Old Egyptian has striking affinities to the Semitic ; but
it is difficult to understand how such a system of writing could have originated
except among those in whose idioms every grammatical suffix was a word of full
and independent meaning, and where the same root or vocable was equally
a noun, a verb, or an adverb. Whatever may be said about its grammar,
the lexicon at least of Old Egyptian fully answers to these requirements ;
and the grammatical character of Chinese and Mexican is well known,
while it may yet turn out that the hieroglyphics of Hamath were borrowed by
the Semitic population from a non-Semitic people, such as I believe the
Hittites to have been. Prima facie evidence is certainly against the assump-
tion that either Semite or Aryan could ever have invented a system of
ideographs.
The inventors of the cuneiform system of writing, at all events, spoke an
agglutinative language. This is one of the most interesting and important
results obtained from the decipherment of the inscriptions, and explains at
once the difficulties and peculiarities which a first view of the Assyrian
syllabary presents. The Assyrians called the agglutinative idiom of their
predecessors Accadian, in distinction from their own Semitic speech;' and a
certain knowledge of Accadian is essential for a right understanding of the
mode of writing which we are engaged upon. Thus the ideograph which
means " a corpse " ^--^ also signifies " to open," two ideas which have nothing
' This seems to me (as to Lenormant, Schrader, and Delitzsch) to result from the correct translation of
a colophon attached to a bilingual (Accadian and Assyrian) vocabulary in IV. A. I. II., 36, i Rl'ik, lines 10 sq.,
where we read : "According to the old tablets and papyri {literally, vegetable of knowledge), the parallel
writings of Assyria and Accad." In W. A. I. III., 55, 2, 9 sq. we have "(The appearances) of the star Curuna
(of the Vine) : (compiled for Esar-haddon) king of multitudes, king of Assyria, son of Sennacherib, king of
multitudes, king of the same Assyria (according to the tablets and papyri), the parallel writings of Assyria, of
Sumir, and of Accad." Here the original Accadian text is not given. In ]V. A. I. III., 64 Rev., 32, we find
"accordmg to the papyri of the tablet, the parallel writings of Babylon." These passages show either that
Accad is opposed to Assyria and Sumir which is placed next to Assyria, or that Assyria is opposed to both
Sumir and Accad. In any case, the first passage contrasts Assyria and Accad ; the last passage proves that
translations from Accadian were made for the ancient library of Babylon after the latter city had passed into
the possession of the Semitic race. As to the question whether Sumir denoted the "Turanian" or the
Semitic population of Chaldea, my behef is that it originally signified the lowland "Turanian" population of
the country which the Accadians found there on their descent from the mountains of Elam ; as this was the
first part of Babylonia to be occupied by the Semitic conquerors, however, the word Sumir afterwards came
to designate the Semitic Babylonians,
3
1 8 LECTURE 11.
in common; and the ideograph which stands for "fortress" ^S^I is
also used in the sense of " death." But the whole mystery is cleared up as
soon as we know that bat in Accadian meant " to open," and " a fortress," as
well as " a corpse" or "death;" and the fact that the same character is
employed indifferently for "corpse" and "open," only shows that the
Accadians had elaborated their method of writing sufficiently to apply the
symbol of some idea to the expression of some other idea which was called by
the same name. But the reason of anything apparently so arbitrary would have
been sought in vain without the key furnished by the old Accadian language.
As we shall see hereafter, the Semites, first in north-western Babylonia,
and afterwards throughout the valleys of the Euphrates and Tigris, took
possession of the country and cities of the Accadians, and gradually extirpated
their language, appropriating their arts and sciences, and above all, their system
of writing. This was developed and improved, just as the Phoenicians, though
not originating the art of writing, took the hieroglyphics of Egypt, and out of
that cumbrous machinery perfected the Kadmeian alphabet. The Eastern
Semite, whose initiation into culture was earlier than that of his western
brother, did not arrive at anything so perfect, greatly as he improved upon
the heritage left him by an alien race : the result of his labours was the
Assyrian syllabary. The principle adopted in the formation of it was this.
The Accadian values of the characters which were significant words in the
old language, were employed as mere phonetic sounds : me, for instance, the
pronunciation of ]* — , no longer represented "a gathering," but simply an
unmeaning syllable. The way, no doubt, had already been prepared by the
Accadians in the case of proper names and a few other words ; but the
extension and consistent carrying out of the principle was reserved for the
Semitic Assyrians. Why they should have kept all, or at least many, of the
numerous values which a single ideograph was able to bear, is far from clear:
it added immensely to the complexity of their writing, and the contrivances
by which the confusion arising from it was sought to be avoided will be
discussed in the next lecture. The most probable explanation is that this
remodelling of the writing was both gradual, and the work of persons who
spoke Accadian as well as Assyrian. Very often the scribes would still be
Accadians ; and to one who was acquainted with the Accadian language, a par-
ticular character, even though ordinarily standing for another phonetic value.
THE SYLLABARY. IQ
might more readily occur as the representative of a certain sound than a different
ideograph which had been already set apart for the purpose. Thus a scribe who
wanted to express the syllable dan or lib might think of m |T before any other
character, and use it accordingly, although it had already been determined
that ^1 It should represent cal in the new syllabary. We must not forget
that the Accadians had themselves begun to use their ideographs as phonetic
characters, so that several of them would be already polyphonous, and the
example thus set, especially if the scribes were Accadians, or were brought up
under Accadian instruction (as we know must have been the case), would
inevitably be followed. To the same cause must be ascribed the retention of
the use of the characters as ideographs. It was found convenient to retain
certain of them as determinative prefixes, one ( j ) to mark an individual,
another (^'^^j to mark a country, and so on, while brevity and rapidity were
aided by an ideographic writing. Hence the hieroglyphic origin of the
characters was never forgotten ; and up to the latest days of the Assyrian
monarchy every character could be used as an ideograph as well as to denote
a phonetic sound. Of course, these ideographs, when they occur in an
Assyrian inscription, have to be read as Assyrian words : i~ , for example, will
not be me, but some grammatical form of the root n'7p " to assemble," or
ramcu, "a herd." We may lay it down as a general rule that the Assyrian
translation of the meaning of an ideograph was never used as a phonetic
value ; that office was left for the Accadian words to fill. It often enough
happens that the phonetic value is met with under a slightly changed form as
the Assyrian rendering of the ideograph ; but this is only because the
Turanian word has been borrowed by the Assyrians and subjected to such
modifications as were needed to make it conform to the structure and
grammar of the Semitic tongues. Thus, "^ nmk, " a building," becomes
muccu; |f^J-| nanga, "a town," becomes nagu; *"'^ ^jj lamma, "colossus,"
lamass'H, etc. I hope to show hereafter how numerous these loan words are,
and what an important testimony they bear, not only to the debt of the Semite
to the Turanian in the matter of civilization, but also to the primitive triliter-
alism of Semitic speech. The general rule, however, which I have just been
stating, admits of one or two exceptions. The sound iz {its, is) was com-
20 LECTURE II.
monly denoted by H, a character which signifies "tree," and is used as a
determinative prefix whenever trees are spoken of. But the Accadian word
for " tree " was ^/s, a vakie which ^j frequently has, iz or rather ets {y}!)
being of Semitic derivation and not employed by the Accadians as a phonetic
power of this character. It is possible, however, that cts {iz) is really borrowed
from the Accadian gis (Semitic p generally answeririg to Accadian g) ; and in
any case its use as a determinative prefix, when it was not pronounced in
reading, had much to do with its pronunciation coming to be regarded as
purely phonetic and non-significant. Nearly all other certain instances of the
Assyrian origin of phonetic powers are to be found in the case of determinatives.
I have noticed above how inevitably a number of different pronunciations
or sounds will attach themselves to an ideograph. All the causes which
bring about their multiplication were at work among the Accadians. Here
was a people well advanced in culture, and whose language, therefore, would
be correspondingly rich, and abound in synonymes. They had, moreover,
elaborated their system of writing, and endeavoured by various contrivances
to make the smallest number of symbols express the largest number of ideas,
as the case of the ideograph for "corpse" and "open" will show; while
dialects in plenty flourished in Chaldea. Berosus says that " a great number
of heterogeneous tribes inhabited Khaldea,'" and at least two Turanian idioms
may be detected in the bilingual tablets. The first idea of writing and the
first hieroglyphics originated among the mountains of Elam before the
Accadai or " Highlanders" had descended into the alluvial plains below, and
Elam up to a late period abounded in dialects. The polyphonous character
of the Assyrian syllabary, therefore, is by no means surprising ; our only
wonder is that it is not greater. As it is, however, the task of learning the
whole of it proved too severe for the ordinary man, and when Assur-bani-pal
wished to give some sort of education to the mass of the people, and enable
the foreigners at his court to read a writing the knowledge of which had
hitherto been confined to the privileged few, he was obliged to have syllabaria
compiled which have done more to give us an insight into the nature of
Assyrian writing than years of patient labour could have done. The king
tell us that " Nebo and Tasmit had made large his ears, and given sight to
his eyes," so that he caused the old learning of Accad, and the syllabaria that
Apiid SynccUi C/i roil icon, p. 28.
THE SYLLABARY.
21
explain it, to be written down and stored "in the midst of the palace for the
inspection of" his " people ;" and the final words, "of my people," are very
noticeable. Trade in Western Asia had long been in a most flourishing
condition ; the merchants of the east and the west met at Carchemish and
Nineveh ; houses were sold and let and money lent at interest while
numberless contract-tablets and other private documents attest that the
necessities of commerce had obliged a considerable part of the population to
acquaint themselves at any rate with " the three R's." Now Aramaic had
become the common language of trade as of diplomacy, and the convenient
Phoenician alphabet was already threatening to supplant the syllabary of
Assyria. To prevent such a humiliation, such a visible symbol that the
sceptre was passing from Assyria to Palestine, it was needful to popularise
the Assyrian system of writing ; and this necessity, as the inscriptions just
quoted inform us, was far more potent than the requirements of the foreigners
from Greece and Lydia, from Egypt and Cyprus, from Arabia and even
India, or than the spirit of an age which resembled that of the Alexandrine
grammarians. The syllabaria, which were drawn up by order of the king,
usually consist of three columns : in the middle is the character to be
explained, while the left hand column gives its phonetic powers, and the right
hand column the Assyrian translation of each of these powers when regarded
as Accadian words. In the right hand column, consequently, the characters
are treated as ideographs, in the left hand column as phonetic symbols so far
as Assyrian is concerned. The careful rendering of each of the Accadian
words (i.e. of Assyrian phonetic powers) is due to the interest felt at this time
in the study of the long-dead language of Chaldea, and to which we owe the
preservation and translation of numberless specimens of Accadian literature.
A syllabary discovered by Mr. Smith when excavating at Kouyundjik
on behalf of the Daily Telegraph, contains a fourth column giving the Assyrian
synonymes of the word by which the ideograph is rendered.' This is
not the first instance of a table of Semitic synonymes ; long lists of these,
with or without Accadian equivalents, and forming a dictionary in the true
sense of the word, are among the treasures of the British Museum, while
' The syllabary is published in the IVth volume of the Cioieiform Ltscriptions of Western Asia, plates
69, 70. In the majority of instances the word given in the third column is an Assyrianised form of the
Accadian word in the first column.
22 LECTURE 11.
other tablets, after setting down a literal translation of the Accadian names of
birds, plants, stones, etc., append the ordinary Assyrian terms in a third column.
With this transition of the syllabary into a dictionary I must conclude the
present lecture. It is not too much to say that the first native lexicon, the
first forerunner of the works of Johnson and of Grimm, arose out of the
complex peculiarities of the Assyrian method of writing. In my next lecture
I shall have to consider the defects of this method, and the devices whereby
the Assyrians in transcribing, and we in deciphering, have endeavoured to
meet and overcome them. This will lead us on to a review of the various
modifications undergone by the syllabary when adopted by foreign neighbours,
until it was finally simplified into an alphabet under the influence of the
practical Aryan mind.
23
LECTURE III.
The Syllabary, continued.
OU will remember my remarking in the last lecture that the Accadian
inventors of the cuneiform system of writing gave proof of their
progress in culture by their attempt to express the largest number of ideas by
the smallest number of symbols. One of the most obvious ways of effecting
this would be by the combination of ideographs ; thus ''papyrus " might be
represented by the hieroglyphics of " writing" and " water," preceded by the
determinative of "vegetable," H t T and "the act of drinking " by putting the
symbol of " water " inside the symbol of " mouth," ^'^-IILT. Such a proceeding
would be suggested and assisted by the agglutinative character of the
language itself, in which the derivatives of inflectional idioms were replaced
by compounds, each member of the compound retaining its full independent
meaning and tone. Thus the idea of "king" was expressed by the
compound xin-gal " great man," and when it was wanted to represent this
idea in writing nothing was easier than to combine the ideographs of
"man" and "great," Hieratic P-fThrnTr^ whence the Archaic tp^Jw^
and Assyrian fc^^ . The plan once adopted was carried out very
extensively, and one of our chief difficulties in the reconstruction of the
ancient Accadian speech is to know when a compound really existed
in the spoken language, or when it did so in the writing alone. We
happen to know that the group of characters ^| || <^ ►^ which literally
signify " house of the land of the corpse," reading e-mad-bat, must be
pronounced arali, and denote "death" or "Hades;" we also happen
to know that the particle " thus," or " if," which is written in characters which
24 LECTURE III.
respectively read su-gar-tur-lal ^ ^ ^^ ]*^ was sounded siigarturlal as well as
tiicundi;' but numberless cases occur in which our present state of knowledge
does not allow us to determine whether the combination existed for the ear as
well as for the eye. Considering the nature of the language, it is necessary,
when any doubt exists, to assume that the compound ideograph really repre-
sents a spoken compound, until the assumption is disproved. The Assyrian
scribes generally regarded these compound ideographs as actual words,
giving a literal translation of them in one column, and the Semitic name of
the object signified in another. For purely Assyrian purposes, however, it
did not matter whether an ideograph were compound or simple ; in either
case the notion it conveyed had to be expressed by a word which did not bear
the slightest relation to the character written down. The most cumbrous
compound ideographs, however, were dropped by the Assyrians ; beyond this
the Accadian system was adopted bodily, though with the important difference
tliat whereas in Accadian these compound ideographs had been phonetic as
well as significant, in Assyrian they became mere signs. Here, then, is one
of the defects connected with the cuneiform method of writing when applied-
to the expression of Assyrian. We have no clue to the Assyrian pronunciation
of a character when used as an ideograph, unless that pronunciation be given
us by the Assyrians themselves ; and our knowledge of the meaning of each
member of a group of ideographs, and, therefore, of the whole group itself,
wdll not aid us in the slightest towards discovering the Assyrian pronunciation
of the group, although we may be thoroughly acquainted with the pro-
nunciation of each separate ideograph. In Accadian, finding that v/;/ " man,"
and gal " great," when combined together meant " king," we may conclude
that " a king " was called migal ; but unless we knew from other sources that
" a king " was termed sarru in Assyrian, our knowledge that nisu was " man,"
and rahu "great" in that language, would not help us towards the discovery
of the fact. The retention of compound ideographs, therefore, increases the
difficulties and labour of the decipherer of Semitic Ass3Tian ; and as this is a
difficulty which naturally would not have been felt by the Assyrians them-
selves, we can only get over it by the patient comparison of variant readings
which may substitute in one text the phonetic reading of an ideograph which
' I accept this on M. Lenormant's authority, but I confess to feeling verj- doubtful myself whether the
four characters in question were ever pronounced otherwise than as ttiaiiuU. The final syllable of the latter
word is the affix da which becomes di after ii, as in Dicngi " the mighty one " for Du/i-ga.
THE SYLLABARY.
25
occurs in another, or by the assiduous examination of the biHn<^ual tablets.
Indeed, without the latter, our knowledge of Assyrian on this side at least
must always have remained extremely imperfect.
But the difficulties presented by the compound ideographs are only an
intensification of those presented by the simple ones. The same idea might
have been denoted by several different words ; and the syllabary of four
columns referred to in the last lecture shows to how great an extent this
was actually the case. The translator accordingly is continually being con-
fronted by the question what Assyrian root out of many possible ones he is
to assign to an ideograph in a particular passage. This difficulty is chiefly
met with in the case of proper names, which are so often written ideo-
graphically ; and one of the main causes of distrust with which the
interpretation of the Assyrian inscriptions was at first received was due to
the uncertainty attached to the reading of the proper names. Outsiders
could not understand how any confidence could be placed in the renderings of
these mysterious legends when the translators differed so materially from one
another, and from themselves at different times, in their transliteration of
royal names, a problem that seemed so much simpler than that of translation,
and, indeed, a necessary preliminary of it. Had it not been for a variant
reading which gave the true phonetic representation, we should never have
known that the second element in the name of Rimmon-nirari, which had
been variously read veiikh, zallus, and likhkhiis T^<^ , was really pronounced
iiirari, " my help," by the Assyrians ; and even now, the name of the Air-god,
whom I have called Rimmon on the authority of certain glosses in the
bilingual tablets, is given as Vul, Ao, and Bin by other scholars." Perhaps
the most provoking case of uncertainty in which this use of ideographs
by the Assyrians has left us is that of the chief personages in the great
Babylonian epic, of which the story of the deluge forms the eleventh
' iMy reading of the name as Rimmon is based upon several reasons, (i) A mythological tablet expressly
interprets the name of the god hy Ravividfw. (2) The original meaning of the name of the god, ^n 1 , is
"wind" or "breath;" from which came the signification of "self," ramanu m Assyrian. Hence in the
bilingual tablets ■*!^'|f- im is generally rendered ramanu. (3) Dr. Schrader has pointed out that the name of
the god is written Ra-man and {Ra-)ina-uu in the name of an eponyme (Canons I., II., W. A. I. II., 68, 2,
21). He has also noticed that the name of another eponyme, Ba7--ku-lid-aii-iii {IV. A. I. III., 47, 3, 8,
compared luith W. A. I. III., 2, 20; 11., 68,2,2,29), proves that the god was sometimes called Barku
"the Lightning" in Assyrian,
26 LECTURE III.
lay. The name of the solar hero around whom the whole epic centres is
written with three characters u Jjla »?- which phonetically would be sounded
Gis, dhii, and bar or Jiias. The first character would be gis, and not iz,
as Mr. Smith reads, since iz was an Assyrian and not an Accadian value.
The name was certainly not pronounced Gisdhubar by the Assyrians ; and
it is "more than doubtful whether it was so even by the Accadians, as the
word seems to mean "body" or "mass of fire," from dhii^ J'l^ "mass,"
and a compound ideograph which literally signifies "wood bound" or
" bundle of faggots," and was used to denote " fire," ^|^y~- '^^^ latter
idea was expressed in Accadian by several different words, and we are no more
warranted in thinking that " fire " was ever called gisbar, than we are in think-
ing that the two characters ^j lij , literally "wood-holding," which are
employed as a synonyme of " king " in allusion to the wooden sceptre he
wielded and used in this sense as a title of Gisdhubar ' himself, were ever
sounded gis-tuk. The name of the Chaldean Noah has been equally a matter
of dispute. Mr. Smith originally conjectured Sisit in reference to the
Sisuthrus of Berosus ; but Sisuthrus apparently turns out to be Khasis-adra.'
Now the Accadian name that appears in the Erech version of the account of
the Deluge which we possess, means " the Sun of life ;" and since zi was
" life" in Accadian, and tarn one of the terms by which the sun was known,
the whole word might easily be read Tam-zi. There are several reasons
which lead to the identification of Tam-zi with the well-known Tammuz ;
among others, the fact that the month Tammuz was called Duzu by the
Assyrians, Du-zi " the son of life " being husband of Istar and a form of
Tam-zi ; ^ and it seems to me, therefore, that we are thus enabled to arrive at
the true Accadian pronunciation of the name. But the very uncertainty in
which it is involved is a good illustration of the obstacle afforded by
ideographic writing to the progress of Assyrian decipherment. It is this
which makes the reading of the astrological tablets so peculiarly difficult.
They were composed at a time when the Semitic settlers in Babylonia were
beginning to learn the arts and sciences of their Accadian neighbours ; the
resolution of the writing of the latter into a series of mere phonetic signs had
not yet been completed, and the ideographs in which the Accadians had
See Note i in the Appendix at the end of the chapter.
* See Note 2 in the Appendix. ^ See Note 3 in the Appendix.
THE SYLLABARY,
27
made their astrologiccil memoranda were reprodueed in full, eked out here
and there by Semitic words and grammatical suffixes. Even the technical
terms which the original inhabitants of Chaldea had written out phonetically
were allowed to remain, though the pronunciation of the corresponding
Assyrian word w^as attached to them: thus ri-ba-an-iia, "conjunction" was
still written, though it was now sounded kas-ri-tu (literally "bond"). The
fact that the same character might stand for several wholly different ideas
adds greatly to the difficulty of these astrological documents ; and no
attempts w^ere made to lessen these difficulties, as in the case of other
inscriptions, since it was wished to confine sacred knowledge of all kinds
to as narrow a circle as possible. Hence these astrological tablets can only be
puzzled out by means of a tolerably wide comparison of passages and a
minute investigation of the bilingual tablets ; and the endeavour I have
made' to give literal translations of the astrological documents, published in
the third volume of Western Asiatic Inscriptions, will, I hope, form a basis for
further work in this interesting direction. When the particular ideographic
force of a character in these tablets has once been determined, the use of
picture-signs renders translation easier than it would be had we to deal with
spelt-out w^ords of uncertain derivation or uncertain application : thus we
know that *"^-Jl^, in which the ideograph of "black" is placed inside the
ideograph of " face" must mean " black face " or " shadow," and *H ] "^j ^£,
" horned," tells us its signification with greater certainty than does its Semitic
equivalent karunu.
This leads me to speak of the advantages resulting from the use of
ideographs which counterbalance its inconveniences. We often find some
Assyrian root of unknown signification interchanging with an ideograph
with the meaning of which we are already acquainted ; and, should the
context suit, w^e are thus enabled to fix the sense of a new word. It is
the compound ideograph, however, in which the superior advantage of
picture over phonetic writing to the decipherer comes most prominently into
view. A large proportion of the ideographs are compound, and it naturally
happens that the ideas represented by compound ideographs are less likely to
be denoted by Assyrian words the forms or special senses of which are
readily to be detected in the cognate idioms ; hence we can not unfrequently
' In the Transactions of the Society of Biblical Anhceology, Vol. III., p. i.
28 LECTURE III.
determine the sense of an obscure Semitic root by observing the nature of the
compound ideograph to which it answers.
But even as regards simple ideographs the Assyrians were not without
means of evading the drawbacks occasioned by the fact that the same
character might be pronounced in more than one fashion. Except in the
case of astrology and kindred subjects, the object of an inscription was to be
read with the greatest ease possible. The 'employment of ideographs was
serviceable to the writer ; but the reader equally demanded that this should
cause him no additional trouble. An inscription was not a puzzle for the
exercise of ingenuity ; and pains had to be taken that mistakes should not be
made in reading it, or time wasted in making it out. So far as the difficulty
of choosing one out of several possible pronunciations of an ideograph was
concerned, the Assyrians felt it as much as we do ; and accordingly, except
where they had to do with very familiar terms, they betook themselves to the
contrivance of phonetic complements which gave the first or last letters of the
word they intended to be read in each particular instance. These phonetic
complements may be divided into two groups, grammatical and lexical. An
ideograph lacked all marks of grammatical relation ; it might be a noun or an
adverb, or a verb of any tense, mood, or person ; where, therefore, the
context did not show clearly beyond the possibility of error to what part of
speech it belonged, a character was either prefixed or affixed, or both, to
indicate how the word was to be read. Thus V^v^T ^CS-?/(i "I conquered,"
l)ut <^ ^ji CISID-ii "acquisition." Generally, when the grammatical indices
were affixed, the syllabic nature of the writing allowed the grammatical
relation to be pointed out as well as the root ; thus ►^^ may be represented
in Assyrian by nadanii "to give," siimu "name," and 5(7/^/// " year ;" but
when we meet with "^ p-SiL we must read IDD-in "he gave" (root j'T^irj),
when *^ ^11 S-uin "name," when *"^*^ ^"^ S-an-na "year," and so on.
It is sometimes difficult, though very rarely so except in the astrological
tablets, to decide whether two or more characters are to be considered
an ideograph with its phonetic complement, or a phonetically spelled-
out word; only the context, for example, can tell us whether >^<^.^TS
is mu-tuv "death," or SANA-tuv "year." The uncertainty here is parallel
to the doubt which arises when a compound ideograph happens to have
exactly the same form as an Assyrian word; Is -^^, we may ask, an
THE SYLLABARY.
29
Accadian term, i.e., a compound ideograph, or is it the oblique case of khisn
" a crown ?" Of course a doubt of this sort can only arise when the
compound ideograph is not amalgamated into one character, but expressed by
a series of characters separated one from the other. Such doubts can only
be settled by a comparison of passages and a practical familiarity with the
inscriptions.
There are certain cases, of no great frequency however, which have to
be carefully distinguished from the use of ideographs. We sometimes
come across abbreviations of common Assyrian words which differ from
ideographs just as in English symbols like viz. or i.e. differ from
abbreviations like K'h. or slid. Such abbreviated words, however, in the
inscriptions are very few, and they are usually found at the end of lines. Ci
sometimes stands for cinii, cinatii " firm," mil for niusii " night," cis for cissu,
cissatii " many," li for livitii " bordering on." I hardly think that the abbre-
viated forms of the prepositions ci for civa, it for ////, etc., belong here : when
we look at the cognate dialects it seems necessary to conclude that the
shortened forms were used in speaking as well as in writing. Such abbre-
viated forms were never employed by the scribe except where there was no
risk of error on the part of the reader, and accordingly they are not one of the
difficulties experienced by the modern decipherer.
I have thus far dwelt upon the difficulties we experience from the use of
simple ideographs: I must now turn to the other side of the picture and point
out that these difficulties are counterbalanced by corresponding advantages. In
the first place, under the form of determinative prefixes, they serve to divide
words and to mark the existence and character of proper names in a sentence.
The upright wedge j denotes that the name of an individual follows, and the
names of women, countries, cities, vegetable substances, stones, grasses,
birds, and animals are respectively preceded by the determinative ideographs,
a list of which will be found in my Elenientary Assyrian Gvaminar. The
ideograph of the plural p** , which may be termed a determinative affix,
is equally useful in showing the number of a noun as well as in marking
its end. This plural affix is often added to the phonetically expressed
plural ending; thus we may have -^^ T<4< TI ^ 5«/r-rt-;// " kings," besides
^\y[ ^f p^ and "^^ p^. The determinative prefixes were, of course, not
usually pronounced ; they appealed to the eye alone, not to the ear. The
30
LECTURE III.
exception to tliis rule would take place only when the gentilic adjective
was written after the ideograph of " country" or " city," instead of the local
name; 'V^ ^^I Hll^ ►^I 1 IT IT ' ^^^ instance, must be read viat Yaliudai
" land of the Jews." The second advantage which we derive from the
employment of ideographs has already been noticed ; it is the clue given to
the signification of an Assyrian word, otherwise unknown, by its being
interchanged with an ideograph of familiar meaning. It is an advantage
which this mode of writing in Assyrian shares with all others of pictorial
origin.
To pass now from the ideographic to the phonetic use of the Assyrian
characters. Here we are met by the existence o,^ polyphones, that great
standing difficulty in the way of decipherment, and the chief cause of the
scepticism with which it was at first received. In the last lecture I have
traced the origin of it, and shown how what seemed an insuperable objection
to the correctness of the key applied to the interpretation of the inscriptions
has turned out to be one of its surest proofs. The variant values which it
was demonstrated the same sign must possess if the system of decipherment
were correct are actually assigned to the sign in the so-called syllabaria^ and
what we now know to have been the origin and primary nature of the whole
cuneiform writing necessitates their existence. But the practical difficulties
caused by their existence remain, and, speaking roughly, can only be got over
by experience. Certain general rules, however, may be laid down for
determining what particular power shall be assigned to a character in a given
instance. These rules, it is true, are not universally valid, and cases will even
arise in which the most practised experience will be at fault. But such cases
are rare, and are getting rarer every day : indeed, we may say that they are
almost entirely confined to proper names and to words of infrequent occurrence,
of which there are no variant readings, or else to those which may be derived
from two different roots of similar signification. Before laying down the rules,
however, it is as well to observe that the student need not trouble himself
about the existence of homophones. Homophones are rare, the Assyrians
having usually dropped a phonetic power belonging to one character which
happened to be the same as one possessed by another. A reference to the
syllabary will show how scanty such cases are, whether the syllables are open
(that is, beginning or ending with a vowel) or closed (that is, with a vowel
THE SYLLABARY, 3 1
between hvo consonants). It is perhaps noticeable that the three chief
vowels, a, i, and //, may all be represented by two characters, JY and ►ij"- for
a, t-^ a.nd J^^ for /, and ^ j 1^ and "^ for // ; but it is probable that when the
second of either of these characters was used, it was intended to mark the
absence of an initial breathing. This at all events was true of ^jj^ and ■^. It
must be remembered that the Assyrians were not composing enigmas ; they
wished their inscriptions to be read ; and accordingly everything was done to
facilitate the reading of them and to remove the difficulties inherent in a
polyphonic system of writing. The chief rules, then, observed by the scribes
(and, therefore, also by the readers) in the choice of one out of many possible
values assignable to a character are these : —
(i) The existence of an ideograph should never be arbitrarily assumed
unless the inscription (like the astrological ones generally) is written through-
out ideographically rather than phonetically. If phonetic complements or
other indications of the presence of an ideograph are wanting, every resource
should be tried before taking the character in an ideographic sense. It was
the neglect of this rule which enabled M. Oppert to get saldha ebus and tsiba
ieris out of saldhac and fsiba'aca, and thus to overlook two important instances
of a remarkable grammatical form.
(2) The triliteral character of the Assyrian language is a sure and
constant guide in the selection of our readings. Quadriliteral roots, mostly
formed by the insertion of r or /, are few in number, and an acquaintance
with them can soon be acquired. When, therefore, we find that only one
value of a particular character will allow of a triliteral root, all other values
leading to quadriliteral or even quinqueliteral ones, we may feel no hesitation
as to the reading to be adopted. Thus when we come across a word
like •^^^fl'^^l we know that the only one of the many values of the
middle character that will give a triliteral word is dan {inii-dan-nin).
(3) The scribes generally gave a clue to the reading by doubling the
consonant, that is, by terminating the preceding syllable or beginning the
following syllable with the initial or final letter of the sound which they desired
should be given to the character used. Not only a dagcshcd letter but a
long or accented syllable also was marked by the repetition of the following
consonant ; and the frequent occurrence, therefore, of these double consonants
is one of our principal helps in determining the reading of a word. Thus
32 LECTURE III.
we know that I^s^ must be read sal-lat "spoil," "^Iq- ^IM] ag-giil-lii
" wheel," and so on,
(4) A knowledge of the grammar and lexicon, as well as of the structure
of the language is indispensable towards settling the reading of a word. Just
as philological reasons must supply the vowels in Hebrew and Phoenician or
the double letters in ^thiopic, so they have often to fix the pronunciation of
a character in Assyrian. The grammatical laws of the language alone will
frequently determine whether a terminal '<^T is to be read as ut or tav or tu,
whether ^^T is to be niiv or nii\ whether ^E is inur, ciji, or khar. But the
lexicon also performs the same function. It is seldom that we have a doubtful
case like H^ *PTTttz ►-^^T ''an oracle," where the second character may be read
either ris or sak, and derived with an equal show of reason from ^i'TD or jl«J ;
the existence of a particular root in Assyrian itself or the cognate dialects
generally settles the question of reading without further trouble. Thus the
existence of the root "i^D "to sell," not only in the other Semitic idioms,
but also in Assyrian, shows us clearly how ni^ 4^ ;/r7;;^-c//r, "saleable thing,"
or " goods," is to be transliterated, even apart from the fact that we once find
a final ri which obliges us to select ciir out of the numerous values of ^, on
pain of breaking our second rule.'
(5) Variant readings of the same passage are a great assistance, more
especially in the case of initial characters. Under the head of variant
readings, we may include variant forms of the same word, one form often
determining the special letter or letters belonging to the root. Thus ''*"^^^|
has been read kJiatstsi and compared with the root nm "to see;" but the
variant '^^jyj{ ^R I i^^'-t^'^ shows that the word is tar-tsi (comp. ^J:J)•
(6) Speaking generally, an open syllable is to be preferred to a closed
one when a character has both powers. Ri, for instance, is to be preferred to
tal when we meet with ^ p| . Usually, whenever a character which was
ordinarily employed as an open syllable was to be read as a closed one, the
fact was pointed out in the way mentioned under the head of the third rule.
(7) Common use had set apart one or two special values for each
character, and unless the action of the other rules interfere, these common
' It was the neglect of this rule that caused Mr. Norris in his Assyrian Dictionary to read the word in
question as " nininiat" which he had considerable difficulty in connecting with " ni/ninairi.^' The character
which he reads niin should be nam.
THE SYLLABARY.
35
and favourite values ought to be read in a doubtful instance. Practical
experience of course can alone decide what these common and favourite values
are ; but a slight acquaintance with the inscriptions will enable us to deter-
mine them in the larger number of cases. Thus tar and cut are more
usual powers of ^^^ than k/iaz or s'il, ►^ is more frequently bat than ///, and
►^1 1 Y more often stands for dan or cal than for lab or rib.
(8) The 8th rule is that a character which represents a syllable
beginning with a vowel is very rarely used after one which terminates in a
consonant, and if an apparent case of this kind occur, the presumption is that
the first character is to be read as an ideograph, the second being its phonetic
complement. Thus *^||^1 ^jllj ^^1 ] 1 is to be re^d ri-e'-iiv " shepherd," and
not ri-bit-uv. This rule, however, admits of exceptions.
(g) The gth and last rule is one that has been of great assistance in
deciphering inscriptions which like those of Assyria do not divide the words
from one another. A word always ends with a line, and a line ends with a
word. Three or four exceptions, at most, can be found to this rule, and even
these occur in the case of proper names like Shalman-eser in a brick legend
lately brought home by Mr. George Smith.'
These, then, are the nine practical rules by which the student may be
guided in his transliteration of the inscriptions. They materially lessen the
difficulties resulting from the use of polyphones, even though they cannot be
said to remove them altogether. But there is a drawback inherent in the
Assyrian syllabary, and quite apart from the polyphony of the characters,
which I have not yet touched upon, although it is really the most serious
defect in the cuneiform system of writing. This system, we have seen, was
originally intended to express the sounds and ideas of a Turanian language,
and its application to a Semitic speech was a later adaptation. Now we all
know how impossible it is to express the phonology of one language in the
alphabet of another. Sounds which are wanting in the one may be fully
developed in the other, while the needs of the scribe may confuse several
distinct letters under one and the same character. All this and more has
happened in the case of the Assyrian syllabary. Accadian was poor in
' It is hardly necessarj' to observe that when two characters (such as ca and ac) come together, the first
of which ends with the same vowel as that with which the second begins, we may infer that they form one
closed syllable (as cac).
34
LECTURE III.
sibilants and dentals. So \\ has to do duty for za (^^T), and tsa (^V)>
and p^Tj for da (^^7)» '^'""^^ '^^^^ (^^)- T^he vocalic ►I^jj was taken to
represent the peculiarly Semitic ayiriy and different shades of sound are
accordin,f;iy confused together in it. But the mischief was least apparent in
open syllables, at least in those which terminated in a vowel. It was in closed
syllables, and in those which ended with a consonant, that the confusion was
greatest. The Accadian made no distinction between the different dentals
and labials at the end of a syllable ; the Semite accordingly who borrowed his
writing had to represent ad, adh and at, ab and ap by one and the same
character. The Accadian blurred the sibilants ; the Assyrian, therefore, had
to use x*^n]p for ^^'^j ^'^^'y ^'■^'i ^^'^ ^^^^- The Accadian m was really mv ; the
Assyrian consequently was forced to use the same character for both in and v.
The Accadian was unacquainted with the sound yu, and so the Assyrian had
to Vv'rite the first and third persons of certain conjugations with the same
character (^ij^), though in the one case it was to be read ''ii (li^), and in
the other yii (v). The consequent uncertainty as to the terminal or initial
consonant of a S3dlable would naturally not press upon the Assyrian, who
would instinctively know what words were required by his language in a
given instance ; but it docs press seriously upon us who are often at a loss as
to the root to which we must assign a particular word. Does the first syllable
of ^^JH ^^p begin with g, c, or k, and does it end with z, s\ is, or s? Mr. Smith
would refer the word to ilDD "to conceal," Dr. Delitzsch to Aram. ^^D^p "wood,"
while my own conviction is that it has the same root as y'^p "jungle."
Variant forms are here almost our sole criterion, and it is only by this means
that we can determine that MH J.Hi is to be read tib, not tip, and referred to the
root ^^12. We shall have to return to this difficulty when dealing with the
phonology of the Assyrian language.
35
APPENDIX TO LECTURE III.
Note i. — Mr. Smith has found a passage in which the name of Gisdhubar
is followed by the syllable ra, which implies that the name ended in r. The
reasons Mr, Smith has given for identifying Gisdhubar. with Nimrod are very
strong; he might have added that the word Nimrod itself may be connected
with Marad, the name of the Babylonian town to which Gisdhubar belonged.
Sir PI. Rawlinson was the first to point out that Gisdhubar was a solar hero,
and that the great Babylonian epic in twelve books, which narrated his
adventures, was based on the passage of the sun through the twelve signs of
the zodiac. The account of the Deluge is introduced as an episode in the
eleventh book of the epic, answering to the zodiacal Aquarius.
Note 2. — I somewhat doubt the reading Khasis-adra. In the first place,
the word or words which Mr. Smith reads thus are found in two passages
only, in the first of which the first two characters (ad-ra) have to be supplied,
and in the second no determinative prefix of a proper name precedes the
group of characters in question. Secondly, in both passages (supposing the
first is restored correctly) we have Adra-kJiasis not Khasis-adra as the Greek
Sisuthrus would require. Thirdly, the name of Tam-zi sometimes has a tiv
affixed, which looks like a phonetic complement indicating the reading of the
name in Assyrian. And fourthly the natural translation of the second passage
mentioned above would be : —
adra khas'is su-na-ta yii-sap-ri-sii.
Then intelligently the dream he caused to explain to him.
4*
36 . LECTURE III. ArrENDIX.
Note 3. — Dii-zi, literally "the son of life," came to signify " the only
son." He was both son and husband of Istar or Astarte, and in one passage
is identified with the Sun-god. It was in pursuit of the dead Du-zi that Istar
descended into Hades. Tam-zi is "the morning sun," that is the sun which
rises again after its nightly disappearance and death, and the similarity of the
words Dii-zi and Tam-zi, both referring to the same deity, seem to have
occasioned the confusion between Du-zu and Tammuz or Adonis in the Semitic
languages. See M. Fr. Lenormant, Sur le nom dc TaminouZy 1876.
^1
LECTURE IV.
The Transmissioji of the Assyrian Syllabary.
MUST now give some account of the modifications undergone by the
Assyrian syllabary in its application to the needs of other languages.
We have already seen that it was itself adapted to the wants of a Semitic
speech from the characters of a Turanian dialect, and just as it was borrowed
by Assyrians and altered to suit their convenience, so it was also borrowed by
neighbouring nations, and more or less changed in the process.
Even before what we may call its Assyrian era, that is before the
Semites had learned the Accadian system of writing, it was used in the great
monarchy of Anzan, or Southern Susiania and its capital Susa. We ought
not to forget that it was from this mountainous country of Elam that the
Accadians had originally descended, and that several facts, such as the use
of papyrus as a writing material, or the ignorance of the palm-tree, go to
show that the syllabary had been invented before their arrival in the fertile
plains of Chaldea/ It was only natural, therefore, that the Susianians should
have employed the cuneiform syllabary from an early date, and that the
characters should be of the Archaic Babylonian form. It must be observed,
however, that they are already cuneiform or wedge-shaped; and this shows
that they must have been derived from the Accadians after the latter had
learnt to stamp them upon clay, and consequently were not a common
heritage which had come down from the period when the Accadians were
still in their early mountain home, Chaldea was so often overrun and
' See Note i in the Appendix at the end of the chapter.
38 LECTURE IV.
conquered by the Elamitcs that there is nothing astonishing in a community
of arts and sciences among them. At present there are but few specimens
of Susianian inscriptions in Europe and these are mostly on broken bricks
from Susa. M. Lenormant has given copies of all that are known in the
second part of his Choix des Textcs Ciinciformes ; and a glance at these will
inform us that the type of character is the same as that found on the bricks
of the primitive Chaldean kings.
Among the four tribes, however, into w^hich Strabo says the country
was divided, that of Anzan or Susa was by far the most civilised ; indeed,
the others, with the exception of the Cassi or Kossaeans, were in a very
backward state. The Amardi, in the north-east, spoke the same dialect as
the aboriginal Turanian population of Media ; and this dialect, the Proto-
medic as it is sometimes called, became of such importance in the Persian
era, when but insignificant remains were left of the Turanian inhabitants
of Anzan and the Magian Medes were exercising a large influence on their
Aryan conquerors, as to be made the representative of the languages spoken
by the Turanian subjects of Persia. It was necessary, therefore, that they
should be provided with a system of writing. This was already in existence
among the Amardians, as may be seen from the inscriptions copied by
Mr. Layard at Mai Amir, but it had been borrowed from the Assyrians of
Nineveh. The Assyrian syllabary had been greatly simplified, polyphones
had been rejected, and only a little more than one hundred characters
retained, including ideographs. Even the forms of the characters had been
made more simple ; thus ^ [so) , has but three wedges instead of four
^^Y^^ Y »-Y^ —
( V )) I'^S l^a-s become ^}^'-] {rak), and ^I>_l is written ^1^
(k/iir). The simplification of form has proceeded to greater length in the
Persian period than in that of the Mai Amir inscriptions, and one or two
characters used at Mai Amir have been dropped. The number of ideographs
employed is very limited, as is also the number of single characters which
express a syllable beginning and ending with a consonant.
The Alarodian nations north of Assyria had similarly borrowed the
Assyrian syllabary in a modified form, though probably not before the ninth
century B.C." The inscriptions copied at Van and its neighbourhood by
' See Note 2 in the Appendix at the end of the chapter.
THE SYLLABARY
39
Schulz in 1828 are inscribed in characters identical with those that occur
on the monuments of Nineveh, except that a double wedge takes the place
of a single line where this passes through another wedge. Thus ^ {pa),
is K^Jc^ , jj^ is t=yy|t^ . The Vannic syllabary, like the Protomedic or
Amardian, did not admit polyphones (except, as it would seem, in one instance
t^J), and used but very few ideographs. As in Amardian, also, characters
that denoted closed syllables were rare ; but on the contrary a great and
extended use was made of the vowels. Thus we find a, u, c, u following one
another, and tar is followed by a, i, c. Dr. Hincks supposed that the final
vowels of a syllable were not sounded at the end of words, par-ri-ni-ni, for
instance, being read parrinin ; but there are no grounds for this view.
We have thus seen what was the cuneiform mode of writing employed
to express the sounds and words of Turanian, Semitic, and Alarodian
languages ; but we must now consider its final and most remarkable adapta-
tion to the wants of Aryan speakers. If we trace the characters from their
pictorial origin through the modifications they underwent in Babylonia, in
Assyria, and in Media, we shall see a constant process of simplification.
But it was among the Aryan Persians that they attained their highest point
of simplification and developed into an alphabet of forty characters, in
accordance with the analytic and simplifying tendency of the Aryan mind.
M. Oppert ' has recently shown us how this alphabet was created. One of
the significations of a character when used as an ideograph was selected and
translated by the corresponding Persian word, and the first letter of the
Persian word was assigned to the character as its value in the new alphabet.
At the same time all the wedges that could be spared were thrown away,
so that the Persian letters are but shadows and maimed relics of the
primitive ideographs. Thus >-^|'^ "time of life," zaya in Persian, is
contracted into >^y< and given the power of z[a, 11); ^\ ''sacrifice,"
liavana in Persian, becomes ^^< and stands for h. In this way, we can
account for the inherent vowels, as they are called, which belonged to the letters
of the Persian alphabet. Thus |c= is used for k before a and i because it
is derived from the Assyrian ^ which signified ** work," karta in Persian,
but ^ IS k when followed by ti, because this character comes from the
Assyrian -^J, the ideograph of the " sun," which was called kuru in
' Jonrjial Asialiquc, Feb. -Mar., 1S74.
40
LECTURE IV.
Persian. Besides the letters of the alphabet, the Persians also admitted a
few ideographs, which again were shortened forms of the old Assyrian
characters. Thus ^/, need not be
pointed out. If this inscription is really of so late an age as it seems to be,
we can easily understand how a native of Babylonia, like Berosus, could
42 LECTURE IV.
have accurately translated into Greek the mythology and history and
astronomy of his country. While Pliny was busied in collecting vague
and contradictory scraps of information about the ancient astronomy of
Chaldea, there were still living men who could have interpreted to him
those very astronomical tablets which have lain so long buried under the
soil. With characteristic contempt for the languages and culture of other
nations, the Romans like the Greeks before them neglected the knowledge
which lay at their doors, and left it to the skill and patience of the nineteenth
century to decipher the records which throw so precious a light on the
history of human civilisation.
43
APPENDIX TO LECTURE IV.
Note i. — As I have pointed out In the Transactions of the Society of Biblical
ArchcEology, VoL L, part 2, pp. 343-45, the ideograph t^ilTry , which is used
to signify "a written tablet," is really composed of two characters, one
denoting ''writing" and the other "water." As the Accadian name of the
ideograph is alal "papyrus," it is plain that papyrus must have been
employed as a writing material while the primitive hieroglyphics out of
which the cuneiform characters arose were still in the course of formation.
M. Oppert long ago remarked that the want of a special ideograph to
represent the palm-tree implied that those primitive hieroglyphics had been
perfected into a system before the Accadians descended into the alluvial
plain of Babylonia. Now it was only after their settlement there that clay
could have been employed for writing purposes ; we are therefore justified
in believing that the use of papyrus preceded the use of clay and that the
cuneiform system of writing in its original shape was employed before its
inventors had left their mountain home. Pliny {Nat. Hist., xiii. 22) states
that the papyrus grew in the Euphrates in the neighbourhood of Babylon
and was there used as writing-paper.
Note 2. — By the Alarodian nations are meant those populations of
Armenia and the neighbouring countries who spoke languages akin to the
modern Georgian. These languages are inflectional in character, but cannot
be connected with those of the Aryan family. The Assyrian mode of writing
seems to have been introduced into Armenia by Lutipri or his son 'Sarduri,
44 LECTURE IV. APPENDIX.
king of the Manni or Minni, the modern Van. As 'Sarduri is probably the
same as 'Seduri, the Armenian monarch with whom Shalmaneser came into
contact in his twenty-seventh campaign (b.c. 832), the date of the introduction
of the cuneiform syllabary into the country is fairly well fixed. The inscrip-
tions copied by Schulz record the names of Lutipri, his son 'Sarduri, and his
grandson Isbuinis I., and then, after a break, contain the annals of Menuas I.
and his four descendants Argistis, 'Sarduri II., Isbuinis II. and Menuas II.
Argistis was the opponent of Sargon, whose inscriptions inform us that he
was preceded on the throne of Van by a king called Ursa, probably an elder
brother. It is not until the eighth century B.C. that we find Aryan tribes
settling in Armenia. Their proper names as found in the Assyrian inscrip-
tions prove them to have belonged to the Iranian section of the Aryan
family and consequently to have formed part of that wave of population
which brought the Aryan Medes into Media and the Aryan Persians into
Susiania.
45
LECTURE V.
Assyrian Phonology.
S the primary object of these lectures is a practical one, I shall confine
myself in this and the following to the general outlines and main
characteristics of Assyrian grammar, reserving points of detail and disputed
questions for a future occasion. For the same reason I shall adopt the
arrangement of an ordinary Aryan grammar, taking phonology, nouns and
pronouns, verbs, particles, and syntax, according to the order with which
Latin and Greek have made us familiar. What has been said on the
subject of the Syllabary will have made it plain that Assyrian phonology is
by no means an easy matter. Where we have a borrowed system of writing
in which the same characters have to do duty for m and v, for tD and *T, for
\ and y, while final consonants are undistinguished, accurate and trust-
worthy decisions upon delicate questions of pronunciation can only be reached
after long and laborious induction. Upon some points, indeed, it is almost
hopeless ever to expect a thoroughly satisfactory conclusion. Such a point
is the Assyrian pronunciation and interchange of m and v. Were these
letters kept distinct in pronunciation and only confused in writing, or had the
Assyrians under the influence of their Accadian neighbours adopted a sound
intermediate between m and v ? In certain cases the latter really seems to
have been the case, especially at the beginning of a word ; but in other
instances such a view is out of the question. Thus, on the one side, b before
the copulative conjunction is assimilated as in eruv-va "he went down, and"
for crnb-va, and here the v or la sound can alone be admitted ; on the other
side, grammatical considerations oblige us to assume m in the case of the
46 LECTURE V,
mimmation ; and the frequent change of the doubtful letter into n before
sibilants, dentals, and gutturals, shows that here again it must have been
m. So far the facts are pretty clear. But now verbs which are yp in
Hebrew are written in Assyrian with the equivocal characters which may be
read either di or v: which of these are we to adopt? Are we to read acmu^
" I burned," amaru " to see," or acvu, avant ? There is much to be said on
both sides ; but when we consider the transcription of Merodach and mana
(inaneh) in Greek and Hebrew, or of maiinuci in the Phoenician bilingual
legends, to say nothing of the numerous roots in which Assyrian in answers
to m in the cognate dialects, or the equivalence of the Assyrian Elamu and
Eliun (fr. nb^) to th'^^, it is better to regard the first pronunciation with m
as more accurately representing the sound of the original. Nothing indeed
can be stronger than dl^ll = nbv where Assyrian must have changed a v (for
ti) into /;/, and the case becomes still clearer when we find ni assimilated to the
following consonant in ^\ovds like ikhkhar "he received," takhkhatsii "battle,"
for the usual imkhar, tamkliatsu, where m has first become n. We shall see
hereafter that one of the distinguishing characteristics of the Babylonian
dialect was its retention of the mimmation. at the end of nouns and verbs ;
and hence when we come to a word like tX^ ^^I ^T cib-nuin-nia " I built,
and " in an inscription of Nebuchadrezzar it is better to assume that the v
or w of the conjunction has been assimilated by the preceding mimmation, or
rather lost altogether, the second m expressing only the accent of the pre-
ceding syllable. This loss of v, through a preceding vocalisation, is of frequent
occurrence in the inscriptions. Like the Hebrew copulative conjunction, the
Assyrian va " and," often appears as il simply, and at the end of a verbal
form may be dropped altogether. Siikaliila, for instance, stands for sukalul-va,
just as dJulbu "good," stands for dlidvdbu (^IlD), the v being first vocalised and
then wholly disappearing. This disappearance of v is more common in
Assyrian than in Babylonian, so that if a case like that mentioned above
occurred on an Assyrian inscription we should rather read abniiv-va than
abnuiiiiii-a, regarding the first v as simply indicating that the accent lay upon
the preceding vowel.
I have dwelt thus long on the changes and interchanges of in and v
because they illustrate so well the difficulties inherent in the subject of
Assyrian phonology, and also meet us upon the very threshold of our
PHONOLOGY. 47
enquiries. We must now turn to the other chief points to be noticed in the
phonetic system of the Assyrians. Let us first take the vowels. A may be
a vowel as well as a consonant, and here again the cuneiform mode of writing
makes no distinction between the two cases. ]} denotes a syllable in
ta-^a-ru " to return," merely a long vowel in kliav-sa-d-nu " forests." Some-
times, however, li ( 4k^->^y ) with or without a ( jy ) is introduced when the
character is intended to be consonantal. The sound originally expressed
by the character has not been exempt from the phonetic decay which has
attacked all the less persistent sounds of Assyrian. The original consonantal
J^ may not only lose its breathing but be still further weakened to i ; and this
weakening of a into i is to be remembered as it forms a characteristic feature
of Assyrian. Before or after u, a is lost altogether. A + 'a or ay a passed into
ai and accordingly is always used for this diphthong in the inscriptions.
Hence aabu must be read aibu {y^}'^) " wicked," " enemy," not aabu as
Mr. Smith gives it in his syllabary opposite the Accadian ^^^^J ►^ITT
" a sorcerer," and the gentilic adjective as we shall see ends in ai not aya,
in full accordance with the habit of the other Semitic languages. I have
already stated that just as ha became a^ so va became u ; but there is one
other observation to be made in regard to this latter vowel. Here again the
deficiencies of the old Accadian script make themselves felt, and the same
character (^111^=) was employed for both u (=1J^) and yu. Thus it comes
about that the first and third persons singular of those conjugations of which
u is a characteristic are expressed by the same character though pronounced
differently. The vow^el i also may be weakened from the consonantal ya.
Thus the third singular of the aorist Kal of verbs is always ispur, though
originally yaspiir. So bitu '' house," is also given as biyatn, where y replaces
a primitive v after its corresponding vowel i. But what distinguishes Assyrian
phonology more especially is the close connection that exists between this
vowel i and 'ay in. The two sounds are repeatedly interchanged, and where
Assyrian weakens ha to /, Babylonian generally has }!. So completely had
a guttural pronunciation of the latter letter been lost that the Assyrian
scribe had to represent the initial of the Heb. HT;/, Gaza, by kh. It was
treated in all respects as a mere vowel, falling away in the Assyrian dialect
after 7C or before a. On the other hand, a reminiscence of its origin was
preserved in Babylonian, and to a slight extent in Assyrian. As just stated,
48 LECTURE V.
where Assyrian has i for ' "to bear." From the same radix we get
lidanu "offspring," with the termination in -dnu on which I dwelt just now.
I must now speak of the cases of the noun ; for Assyrian like Arabic
preserved the three cases which were originally possessed by all the Semitic
languages. These cases ended in -w for the nominative, -i for the genitive,
and -a for the accusative, and had once been kept clearly distinct. At the
PHONOLOGY. 55
epoch, however, to which our earhest Assyrian monuments mount back
this clear consciousness of their distinction had been lost and a tendency
had set in to use these cases one for the other, the accusative and genitive
being employed for the nominative and vice versa. The later the age
of the inscription, the more frequent does this misuse of the cases become,
although to the last the distinction between them was never wholly
forgotten, and accurate writers like the scribes of Assur-bani-pal took care
not to confuse them. The vowel-ending -i came to have the preponderance,
partly through its being a weakened form of both a and u, partly through
the influence of the prevailing plural termination in -i. We learn from
comparative philology that all these case-endings were originally long
although the later inscriptions almost invariably make them short ; and
that there was a time when only the objective termination in -a existed,
-i and -il having been subsequently developed out of it. Here is another
illustration of the lesson taught us by the science of language that the
object historically precedes the subject, the objective case the subjective.
But the Assyrian case-endings were distinguished by a further peculiarity,
which has been lost by all the other Semitic languages with the exception of
one dialect of the Himyaritic inscriptions. This peculiarity is a terminal m,
which closed in the vowel, and was probably pronounced much in the same
way as that final m in Latin which could be elided before a vowel. Instead
of m, Arabic has 11 in the same place, and as this nasal termination goes by
the name of nunnation, the similar phenomenon in Assyrian has been termed
mimmation. M was older than n since we elsewhere find final m in the
Semitic tongues becoming n, in the sign of the plural for instance, just as in
Greek a terminal m changes into the nasal. The mimmation came to be
more and more omitted in Assyrian, and we often come across inscriptions in
which it is absent altogether, while in others its use is extremely irregular.
The southern dialect of Babylonia was far more conservative than the
northern dialect of Assyria in this particular respect. Up to the last the
Babylonian inscriptions are characterised by a frequent employment of the
mimmation, and while the presence of it is exceptional in Assyrian, the
omission of it is exceptional in Babylonian. As we shall see, the mimmation
is found in verbs as well as in nouns, and its existence explains the primitive
length of the case-vowels.
56 LECTURE V.
As in the other Semitic idioms, the genitive relation is ordinarily denoted
by that close connection of the governed and governing words which allowed
them both to be pronounced like a sort of compound in one breath, and
shortened the form of the first or governing noun. This shortening is effected
in Assyrian in a very simple way, by dropping the case-endings and mim-
mation of the first word, and attaching the genitive termination in -i to the
second. Thus from zicini (Heb. zecer) or zicirum " memory," we get zicir sumi
" memory of the name," zicir sarruti " memory of the sovereignty." Surd roots
lose the last syllable and with that the last radical, as ^ar from darru. The
shortened form is the one invariably assumed by the construct genitive, so
that whenever we meet with cases like belutu Assur "lordship of Assur,"
we may be quite sure that we have to do with two nouns not in the
genitive relation but in apposition to one another, so that the literal render-
ing of the clause would be "the lordship, viz., Assur." Besides this con-
struct genitive, the Assyrians made increasing use of a periphrastic genitive,
with the relative pronoun sa "which" placed between the first noun with the
case-ending of the nominative and the second noun with the case-ending of
the genitive. Thus instead of zicir sumi we might have zicirn sa sumi, literally
" the memorial which (belongs to) the name," and so " the memorial of the
name," sa coming by degrees to have the force of a mere preposition " of,"
like the Greek ^apLv " because of."
There were but two genders in Assyrian, masculine and feminine,
abstracts being included under the latter. Many feminine substantives
have no distinctive termination, and their gender can only be known from
their meaning, plural form, or employment with feminine adjectives, like
htmmu " mother," iizmi " ear," lisanu " tongue." Those that have a distinctive
suffix are of three kinds, (i) Feminine abstracts in uttL already mentioned.
(2) The general feminine ending in -atu (e.g., napsatu, belat, contracted
beltu, pulkhatu, pulukhtu.) (3) In -Uu weakened from -atu (thus both
belatu and belitu; binitu and bintu). The latter was the only form admitted
in roots ]J"^,
Assyrian possessed a dual as well as a plural, though its use was confined
to a few words which denoted pairs like uznd "the ears," katd "the hands,"
sepd "the feet." These examples will show that it was expressed by the
termination a. The plural was formed in several ways. The oldest was that
PHONOLOGY. 57
in -dnu, etc., which was used indiscriminately for both masculine and feminine
substantives (e.g., feminines emukdnu, '' powers " risdnu "heads.") Some-
times, however, we find feminine nouns which have not only this but also the
more modern plural termination in -dtu, as e.g., pdmi or pdtu " faces." Some-
times, again, masculines which have adopted the later masculine ending are
occasionally met with under the form -dnu. Instead of -dnu, the weakened
form dni is usually found apparently through the influence of the common
plural in -/ (e.g., kharsdni "forests," duppdni, "tablets"). Just as the original
case-ending -a was softened into u, so besides -dnu there is a rare form -unu
which occurs in the word dilunu " buckets," and a few others.
But besides dilunu, there was another plural diliitu, which furnishes us
with an instance of a very common Assyrian plural in -iitu. It was set
apart for masculine nouns, though the fact that a precisely similar formation
denoted feminine abstracts singular shows that it was properly and primarily
the characteristic of feminine nouns. How it came to be restricted to the
opposite gender is an example of a phenomenon that frequently occurs in the
history of language. It was the form of the plural adopted for all adjectives
and present participles, as well as for roots ii"'?, etc.
The ordinary plural of masculine substantives was, however, one in -i
or -e. This resembles the construct masculine plural in Hebrew, and like the
latter has lost a final m or n. But whereas Hebrew confined this contracted
form to the construct genitive, Assyrian applied it to all cases alike, thus
giving another illustration of its liability to phonetic decay. The result of
this was that in many nouns it is impossible to distinguish between the
plural and the genitive singular, both being pronounced alike, though an
attempt was sometimes made, especially in Babylonian, to keep them apart
in writing by using e instead of i for the plural. This was the easier from the
fact that the vowel of the plural was when correctly sounded longer than
that of the genitive singular, the long vowel of the latter having been worn
away before the action of decay had begun to break down the plural ter-
mination. Another mode of distinguishing between the two forms was
adopted in the case of dissyllables, when the accent was on the first syllable
and the second syllable was short, by dropping the vowel of the second
syllable in the singular, and retaining it in the plural, of course accenting it
at the same time. Thus ndcri is " enemy," (genitive singular), naciri
^8 LECTURE V.
"enemies," nakJili "valley," nakliaUi (for nakhdli) "valleys." But in other
words the confusing consequences of phonetic decay are as conspicuous as
in our own "sheep," and the context or grammar alone can determine whether
the word is singular or plural. When we meet, for instance, with a sentence
like rabbi bitu we may know at once that rabbi here must be the plural, since
the law which regulates the construct genitive would require rah bitu, were it in
the singular. It may be remarked that besides this plural in -e many nouns
also retain the earlier plural in -dnu (e.g., sarrdnn by the side of sarri).
In opposition to the masculine plural in -i there is the feminine plural
in -dtii {-dti, -dta), which is sometimes weakened to -Uu or etu, especially in
the case of adjectives used as substantives (e.g., esreti "sacred places,"
perhaps connected with the Hebrew ''asherah). Many substantives are of
common gender, and therefore admitted of both the masculine and
feminine plural, like babu " gate," which has the two plurals babi and
babdtu. This fact will suggest a solution of the curious phenomenon I
alluded to just now, the restriction, namely, of what was originally the
feminine plural to masculine nouns.
The Assyrian cardinal numerals, as in the other Semitic languages, have
two forms, one feminine and the other masculine, but those from 3 to 10 use
the masculine with feminine nouns and the feminine with masculine nouns.
This, again, is another illustration of the transition of meaning in what was
primarily the feminine form of plural nouns. In old Babylonian, however,
we find traces of a different and more correct usage, where besides ciprdtu
irbdi "the four zones," we have tupukdtu irbittu with the same meaning.
The larger number of the cardinals are met with on the monuments ;
only those in brackets in my Elementary Assyrian Grammar are still un-
detected. Their conjectural restoration, however, is pretty certain. Instead
of writing each number in full, the Assyrians generally made use of a
symbolical mode of expressing them like our ciphers. In this system,
an upright wedge (J) denoted i, two upright wedges (JJ) 2, and so on.
For 4, instead of writing the wedges one after the other, three were written
in one line and one beneath, thus ^ ; and the same arrangement was
adopted as far as 7, when a third line was added (^). For 9, there was
besides ^\ the abbreviation ^. Ten was expressed by ^, and the
succeeding numerals were denoted by the help of this arrow-head and the
PHONOLOGY.
59
wedge. Thus ii was ^, 15 ^, 20 ^^, and so on. With 60, however,
a new system begins which is at first sight somewhat puzzHng. The
Accadians had attained to remarkable mathematical proficiency, and had
found that the duodecimal was scientifically a more convenient numerical
system than the decimal, the only recommendation of which is that it is the
first suggested to the savage by the fact of his possessing ten fingers. They
consequently made 60 their unit, and accordingly in the development of their
arithmetical symbols represented both i and 60 in the same way by the
upright wedge. It is sometimes difficult, therefore, to determine whether
I or 60 is intended in the inscriptions; indeed this can frequently be done
only by the help of the context or internal probability. The wedge for 60,
however, was generally thicker and larger than that for i. After 60 there
is no difficulty, since the combinations fy, fyy, 61, 62, etc., and f^,
YKK 7^j ^^' ^^^-j ^^^ ^^^ otherwise met with. For 100 the character ]*^
me was employed, since me in Accadian signified " multitude," and then " one
hundred," in which sense it was borrowed by the Semites. 1000 was easily
expressed by prefixing ^ (10) to the sign for 100.
A noun in the masculine plural always follows the cardinals, as esritu alpi
" 10 oxen;" in the case of weights and measures, arithmetical terms, etc.,
however, the noun is put in the singular, thus esri mana " 10 manehs." In
the latter instance the measure is often preceded by the preposition ina " by,"
followed by the sign of unity, but without any change of meaning : cc ina
I. ammat (t^llJ^ \*^)y ^i" example, being literally " 200 x i (i.e., 200) cubits."
The ordinals were formed from the cardinals, with the exception of
the terms for "first" ristdnu, from risu "head," and makhru "foremost."
Thus "second" was sannu, feminine sannutu (with nn for ny, i.e., nw),
"third" was salsii, feminine salistu. A formation similar to ristanu denoted
relations of time, thus saniy-dnu was "the second time," salsi-y-dim "third
time," etc. Collectives took the form sunnu "a pair," plural simne, sulsu " a
triplet," etc.
The names of the fractions, on the other hand, seem to have been
derived from the Accadians, although the forms just mentioned, sitlsu, rubu,
etc., appear to have been sometimes applied to this purpose. Thus sussdmi
was "a third," from the Accadian sussana, sinibu " two-thirds," from Accadian
sanabi, parapu " five-sixths," barn or mdsu " one-half." This is another
6o LECTURE V.
illustration of the borrowed nature of the mathematical knowledge of the
Assyrians, who had to take even their name for the mathematical unit, the
soss (60), from Accadian. So too saros (3600) and ney (600, symbolised by Y^
which also equals 70) are of Accadian origin.
Indeed it will be found on closer inspection that most of the names of
weights and measures came like the syllabary from the Turanian predecessors
of the Semitic population in Western Asia. Thus the standard of length,
the ca^bn (7 miles), pi. caslbume, is an Accadian word derived from ka^ "two,"
and bu " length," and from being a measure of length it came to be used as a
measure of time. In this case two of our hours went to make up i cas'bu,
and some of the astronomical reports, sent in to the king from the observatories
of Assyria, mentioned that at the vernal equinox day and night were equal
{sitkulu), there being 6 ca^bu of the day and 6 of the night. Other astro-
nomical terms had a similar Accadian origin, since this science also, together
with the formation of a Calendar and the division of the year, owed its
beginning to that ancient people. From them, too, came another word,
which after being applied to the valuation of money, and passing through the
vocabularies of Greece and Rome, is still preserved in our dictionary. This
is the maund or maneh, the /xm of the Greeks, the mina of the Romans, which
appears in its most primitive form mana in the Accadian column of an ancient
table of laws. Already at that remote date it was employed to measure
the precious metals, and so would be readily taken up by the Semites, those
merchants of the old world. Through them it made its way to Egypt and at
a later period to the nations of Europe, and still remains an enduring
monument of the debt which civilisation owes to the forgotten thinkers of
Chaldea.
6i
APPENDIX TO LECTURE V.
Note i. — ^The land of Kaldu or Caldu is first mentioned by Assur-
natsir-pal (col. iii. 24.) in B.C. 878, and in b.c. 850 his son Shalmaneser
speaks of the district as lying below Babylonia on the Persian Gulf (Obelisk
83). It was not till a later period that the Caldai occupied Babylonia, and
under Merodach-Baladan made themselves so important and integral a part
of its population as to give their names to the whole country. The word
Casdim is best explained by the Assyrian root casadu " to possess," or
"conquer," so that the Casdim will be those Semitic "conquerors" who first
settled in Sumir or Shinar, and finally succeeded in extirpating the power and
the language of their Accadian predecessors.
Note 2. — The following exceptions to the general rule of Assyrian
accentuation may be noticed :
(i) The enclitic -va threw back its accent upon the preceding syllable,
as remarked in the text.
(2) The possessive pronouns of the first, second, and third persons
when suffixed to a noun threw the accent back upon the preceding syllable,
3iS paml-ca " thy face," ramanu-su " himself," ramanu-sun " themselves."
(3) The possessive pronoun suffixes of the verb, with the exception of
the second plural and third masculine singular, threw the accent back upon
the preceding syllable, as radib-d-ni "pierce me," itticriih-d-ni "they were
estranged from me," htcassipi-ni " thou (fem.) didst reveal to me," pitd-si
" open for her." A double accent is even permitted in icsudd-su-va " he
conquered him, and."
(4) The vowel between the first and second radical was accented in the
present of Kal, as isdcin, isdcimi. So, too, in the quadriliteral ipardsid.
(5) The penult was accented in the third plural masculine, and .perhaps
62 LECTURE V. APPENDIX.
also feminine, as itsbntn " they seized," immdru " they were visible," ituru
" they returned."
(6) The penult was accented in the present Kal of verbs ]^"h, as iseri
from ^ID and isUi from ]^0^.
(7) The third person singular of the subjective aorist of Kal and Niphal
was accented on the penult, as ippisidu " it was alleged," inukhu " it had
rested."
(8) Dissyllabic nouns whose first syllable was accented, the second
syllable being short, accented the second syllable in the plural, as nakhdli
"valleys," nac'iri "enemies," in contradistinction to the genitive singular
ndkhli and ndcri.
(9) Certain nouns accented the penult, like agurn " cement," citstsilu
"royal," barzilu "iron," cidinu "ordinance," cuduru "landmark," addnu
" season," siih'imn " alliance."
It will be seen from this that the accentuation of Assyrian words agrees
very remarkably in many particulars with that of Ethiopic as described by
Dr. Trumpp in the Zeitschrift der Deutschen Morgenldndischen Gesellschaft,
Vol. xxviii. 4 (1874). See my paper on " The Tenses of the Assyrian Verb,"
in the Journal of the Royal Asiatic Society, Vol. viii., p. 19 (1876).
63
LECTURE VI,
The Pronoun.
HIS evening I come to the pronouns, one of the easiest parts of Assyrian
grammar. A definite number of words have to be learned by heart,
and there is no need of carrying certain type-forms in the head to be appHed
to some strange word we may meet in the inscriptions, as is the case with the
nouns and verbs. You will observe how closely the Assyrian pronouns
resemble those of the other cognate Semitic languages, while some of them
present us with older forms which have been lost in the other dialects.
A list of the personal pronouns will be found in my Elementary Grammar.,
p. 57. You will notice that the second syllable of the pronoun of the first
person plural is given doubtfully. The word unfortunately is only once
found, in the Babylonian text of the famous Behistun inscription (1. 3), and
there the reading is uncertain. The first and third characters are clear
enough, but the middle one is doubtful. According to Sir H. Rawlinson's
cast it looks like ^^ ga; but the analogy of the cognate dialects would
lead us to supply some character with the value of nakh. You will also notice
that both the first and second persons singular have a double form. The
first form anacu and atta or atti^ is the usual one, and the one, too, which
agrees with the ordinary first and second personal pronouns of the other
Semitic dialects. Ydti and cdtii have a more substantival force, like our
''myself," "thyself," and are accordingly generally found at the beginning
of a sentence. Thus we read (W.A.I. I., 68"; II. 19, 21,) ydti Nabunahid
" as for me Nabonidus ;" ca-a-tiv a-mat-ka man-nu i-lam-mad " as for thee, who
learns thy will." Cdta (the accusative oi cdtu) may, however, be used for the
64 LECTURE VI.
sake of emphasis after a preceding verbal-suffix -ca as (Sm. A. 183) iisamkhar-
ca cdta " I cause thee, even thee, to be present." It is very interesting as
affording us an example of the guttural form of the second personal pronoun
used otherwise than as a verbal-sufhx, and may be a relic of a very old
pronoun which has been lost in the cognate Semitic idioms. The termination
both of cdtu or cCita and of ydti is the same as that of the feminine in nouns,
and of the longer form (sunutu, etc.) of the third personal pronoun plural.
It will be remembered that the Semitic languages express an abstract
idea by a feminine suffix, and in this way we can explain how the feminine
termination came to be attached to the personal pronouns when they were
employed in an abstract sense ("myself," "thyself," "themselves"). We
never meet with the nominative ydtu or the accusative ydta in the case
of the first personal pronoun, but only with the lighter case-ending ydti.
There was something in this latter ending which suited the ears and vocal
organs of the Assyrians ; tii and ta had always a tendency to become ti, and
in some substantives like tucidti it is the form almost invariably used. Ydti
(Ci^y^r If >^y<) was frequently shortened into yati, and in one instance
(Sm. A. 225, 55.) we find aiti, according to the Assyrian phonetic law by which
a -^ a represents not only aya but also yd and ai. It is interesting that traces
of the same pronoun are to be found in the Hebrew 'an-i for an-ya, which
denotes the first person as well as the more usual 'anochi, and in the Ethiopic
ana " I," for an-{y)a. Besides ydti, the longer, more emphatic form yatimd
is met with in the inscriptions, which literally signifies " myself here," ma
being the suffixed demonstrative pronoun. We shall meet further on with
numerous examples of this suffixed pronoun. Instead of ydti, however, we
sometimes find ya-a-si (ydsi), which occurs more generally in the middle than
at the beginning of a sentence. Thus Sm. A. 108, 3, salini-mu ydsi libba-cunu
"greeting (from) me [as far as I am concerned] to you [your hearts];" 225,
35, sa naciru sanavva eli ydsi kats'u la nbiliL ina lihbi "which no other
enemy except me (with) his hand has touched [brought his hand into the
midst (of it)]." Elsewhere we have aisi just as we have aiti by the side of
ydti. Now although there is no doubt as to the meaning of this word ydsi as
it is used in the inscriptions, and of its being a synonyme of ydti, the origin
and real signification of it is extremely hard to settle. Prof. Schrader
believes that the termination si is a suffix like ti; and though he does not go
THE PRONOUN. 65
on to say so, I suppose he would connect the suffix with the pronoun su, so
that ydsi (or ydsu) would primarily have meant '' that me," or " me namely."
He has support for this view in the fact that by the side of sinati *' them,"
and the sufhxed possessive pronoun of the second person plural cimu we seem to
get also the very rare forms sinasi and cunusi. Thus in the old Babylonian in-
scription of Khammuragas at the Louvre, which is one of the oldest inscriptions
written in Semitic Assyrian that we possess, and which is certainly earlier than
the sixteenth century B.C., we read (ii. 6.) lu as-cu-un-si-na-si-iv " I made them,"
and in Sm. A. 108, 4, lu dhub-cu-nu-si mi-nav-va "your good deeds are
numbered, and." But for my own part I cannot assent to his theory. Con-
sidering the similarity that exists between the two characters ^*- si and
>~) Ittanaphal, ittanacatum
{lb) Iphtanael, yuctanattum
{4b) Istanaphal, yustanactinn
(5a) Itaphal, yutadhih
{dfC) Shaphael, yuscattum
no
LECTURE VII. APPENDIX.
(3^/) Pael Passive, ytuuttum
aitUi7)i)
(5a in Assyrian.
IP\ A. I. I., 9, 14.
ASSYRIAN SYNTAX. II7
of the final at into all. The verbs, however, show that a sense of the
distinction between the terminations of the masculine and feminine had
become weakened as far back as the twelfth century B.C. Tiglath-Pileser I.
says of ''the great gods" Hi rabi, who are, of course, masculine, aga'a tsira
tiippira-SH "the supreme crown ye have entrusted to him," where the verb is
feminine; though, it is true, we might explain the form tiippira here as belonging
to the objective aorist, and so standing for tuppiru'a {tuppiruna). But the
frequent use of a feminine nominative with a third person masculine in the
later days of the Assyrian empire favours the first view. Assur-bani-pal tells
us that ** Istar," ana ummani-ya siitta yusapri-va ciham icbi-sunut, ''to my
soldiers a dream disclosed and thus declared unto them" (Sm. A., 221,
23), and in the " Descent of Istar into Hades " (I. 3) we have Istar banat 'Sini
uzun-sa .... iscim " Istar, the daughter of the moon-god, set her ear."' In
the Assyrian translation of the famous tablet of ancient Accadian laws
{W. A. I. II., 10) there is a very curious neglect of the genders, the masculine
being used for the feminine not only in the case of the verbs (as ictabi for
tactabi) , but even in that of the pronouns (as atta for atti and su for sa or si) ;
but the explanation of this must be sought in the fact that the translation was
probably made by a scribe of Accadian origin, who had not been accustomed
to distinguish the two genders from each other. Nisi " men," however, in the
abstract sense of " humanity," is sometimes used with a feminine adjective, as
in W. A. I. III., 41, 39, where we find nisi disdti " abandoned men," just as in
the Behistun inscription (line 16) 7^/^ww " people " is joined with the plural
gabbi "all," and itticru'' "they revolted." The want of agreement in number
is much rarer than that in gender, and occurs only with nouns used in a
collective sense, or where the adjective is more or less independent of its
substantive. Thus we not unfrequently find the expression ^arri alic makhriya
[W. A. I. II., 21, 29) " the kings who went before me," where the full force of
the words would be "the kings, each goer before me." The decay of the dual
naturally caused it to be generally construed with a plural predicate.
The attribute, we have seen, may be the genitive of a second substantive
as well as an adjective. The genitive-ending of a large class of Sanskrit and
' Compare, too, in the Babylonian transcription of the Naksh-i-Rustam inscription (9, 10), the use of the
iermmrie matati " countries," with the masculine inassunu ''will bring," which is similarly found with the
masculine ittnrunii "they have returned" at Behistun (7).
Il8 LECTURE VIII.
classical nouns was originally adjectival, the Greek SrjiJiov, for instance, which
stands for an earlier brjixoaio (Sansk. -asya), being formed by the adjectival
suffix tya, and so differing but little from StjfjLoa-LOf- Like the attributive
adjective, the attributive genitive followed its substantive in the Semitic
languages, and the two words were as closely combined in pronunciation as
they were in sense. They were in fact pronounced in one breath, an external
symbol of the fact that they together made up but one idea. As we have seen
in a former Lecture the first or governing noun lost its case-endings in
Assyrian when in the "construct state," so that "the house of the king"
would be bit darri instead of bitu ^arri. Originally the attributive noun took
the light genitive ending in -z, which was a weakening of the termination of
the objective case (-rt), just as the genitive relation itself was a weakened
form of the objective relation. Two substantives might follow one another,
each in the construct state, as in the other Semitic languages ; thus we get
Nabii pdkid cissat same ti irtsitiv " Nebo, the overseer of the hosts of heaven
and earth " {W.A.I. L, 51, i, 13). As also in the other Semitic languages,
however, the possessive pronoun suffix might be regarded as an adjective, and
attached to the second member of a genitival compound without causing the
latter to lose its case-endings. Unut takhazi-sunu, for example, is "their
munition of war " (Lay. 16, 46), nisi takhazi-ya " my men of war " {W . A. I. L,
39, 44). A violation of the rule by which the governing noun lost its case-
endings is a very rare occurrence, and is usually capable of explanation.
Sometimes the nouns are in apposition one to another, and not in the genitive
relation, as in such a phrase as belutn Assur "the lordship of Assyria"
(properly "the lordship, that is Assyria"), or pulkhit melain Assiivi "fear, even
the onset of Assyria," that is, "the fearful onset of Assyria."' Sometimes
the first noun is plural, and the final vowel is consequently a mark not of case
but of number, as in rabbi biti "chiefs of the house." Sometimes the anomaly is
due to the fact that the first noun forms part of a compound preposition, like
ina hiculti "in the service of," ina libbi "in the midst of," where the analogy
of the other prepositions itti "with," arci "after," eli "above," adi "up to,"
etc. (like the Heb. v!?, Hi?, etc.) has been followed. One of the few real
exceptions to the ordinary rule is to be discovered in W. A. I. II., 66, 4,
where Beltis is called bucurti Annv " the eldest-born of Anu," instead of the
■ Mclam I derive from the root ni?.
ASSYRIAN SYNTAX. Iig
more correct biicrat. In fact, so far as I know, the only cases in which the
rule is broken are where the feminine ending -ati with two short vowels follows
a preceding short vowel.
I must here turn aside for a moment to notice an apparently converse
anomaly that occasionally presents itself in the texts, where a noun, though
not in the construct state, seems to be without the case-endings. The
anomaly, however, is apparent only, and arises from the deficiencies of the
Assyrian mode of writing. The last root-syllable of a word, if it begins and
ends with a consonant, may be expressed by a single character ; in this case
the short vowel of the case is not added but left to be supplied by the reader.
Thus kakkadu " head " may be written ^ ^J, where we must read not
kak-kad but kak-kadu ; panu lijunii "evil face," (\^ <|>- >^, where the first
ideograph must be sounded panu and not pan.
Nowhere can the attributive sense of the genitive relation be seen more
clearly than in the way in which abstracts were frequently represented in
Assyrian as in the cognate languages. This was by combining a substantive
expressing the possessor or subject with another substantive expressing the
attribute, so that bel-khidhdhi "the lord of sin" would mean "a rebel"
(W. A. I. I., 2>7, 39), ^lis riicubi "the man of chariots," "the charioteers," and
er sarriiti-su "the city of his royalty," "his royal city." So close is the
combination that when the plural is required the sign of it may be attached
to the second substantive only, while a negative sense may be obtained by
prefixing the negative particle, as in la-bel-cits's'u "the not-lord-of-the-throne,"
that is "an usurper." We all remember similar modes of expression in the
Bible, where " the daughter of Zion " signifies " the inhabitants of Jerusalem,"
and the New Testament writer uses " the son of peace " instead of the abstract
" peace," in imitation of the Hebrew idiom.
As in the other Semitic languages, so, too, in Assyrian the construct state
came in later times to be replaced by an analytical periphrasis. The synthesis
between the two parts of the idea, the subject and its attribute, was broken
up, and a construction adopted which involved an assertion of judgment. In
other words, the predicate took the place of the attribute. This was effected
by placing the demonstrative sa " that," which in course of time assumed a
relative signification, between the substantive and its attributive genitive.
'Sarru sa Assur "the king, that (is) Assyria " was substituted for the simpler
I20 LECTURE VIII.
^ar Assur "Assyria's king," and so the pronoun sa came gradually to have the
force of our preposition " of." The analytical character of this construction
can best be observed where the predicate precedes the subject, as in sa sanati
arkhi ''of the year the months" {W. A. I. III., 52, 43), though in such
instances the second noun ought to have the possessive pronoun suffix, as
insa Cambuziya aga-su akhu-sii "of Cambyscs this man his brother," a con-
struction permissible in Ethiopic, but rarely found in Arabic and later
Hebrew. In some cases the introductory sa would be most idiomatically
rendered by "as to" or "regarding" (e.g., sa Ambarissi malic-simu "as
regards Ambaris their king"), and we even find instances in which it is
dropped altogether.
The periphrastic genitive served to express the superlative as in the
phrase Akhiiramazda rabii sa ili^ " Ormazd, the greatest of the gods."
The predicative relation constitutes the germ of the sentence. But the
sentence does not become complete until it possesses an object. A subject
implies an object as much as it implies a predicate, and a predicate can only
be made definite and concrete by being provided with an object. " The king
is a conqueror " is a merely general statement ; it becomes definite by the
addition of the object he has conquered. Predicative sentences have little
practical utility; the assertion that "man is mortal" may be very fitting
for works on logic, but it would not be of the slightest assistance towards
supplying us with the needs of our every-day life. The first speakers must
have contented themselves with objective sentences, and left predicative sen-
tences to their more intellectual descendants. Indeed, since the satisfaction of
his wants must have been the primary motive in primitive man for the
creation of language, the indication of the object would have been of supreme
importance to him, and accordingly we find in all languages that the objective
case is older than the subjective.
Now an objective sentence is one in which the action passes on from the
subject to the object ; where, therefore, the two factors are contrasted together
in the mind as independent but related. The object is conceived as essentially
distinct from the subject, unlike the predicate which is conceived as forming
part of it. The most natural way of contrasting the subject and the object
would consist in their immediate juxtaposition, and the primitive savage who
said, " Me, pear!" would be quite as intelligible as the Englishman of to-day
ASSYRIAN SYNTAX. 121
who says, " I want a pear." We find children constantly returning to this
primitive method of expression. If the subject and object are thus contrasted
by being placed side by side, the verb or representative of action will have to
follow the object, that is will come at the end of the sentence. And such we
find to be the case with barbarous languages which have arrived at the con-
ception of a verb, as well as with more cultured ones which have retained
their original usage. This, too, is the order of the ideas and words in a
sentence adopted by the deaf and dumb. In course of time, however, the
natural order is likely to be replaced by the logical order, according to which
the object will follow the verb or representative of action. That to which the
action passes, and where, therefore, it finds its rest and fulfilment, is logically
last.
Unlike Hebrew, Assyrian observes the natural order of words in the
sentence rather than the logical one. The Assyrian verb regularly comes at
the end of a clause. The chief exception is when the objective form of the
tenses is used, and here, in unconscious remembrance, it would seem, of the
time when the objective aorist and present were merely verbal nouns, the
verb precedes the case it governs. The personal pronouns, again, are affixed
to the verbal forms, but this is due to the fact that the verbal forms were once
nouns, to which the possessive pronouns were attached according to rule.
When the possessive form of the pronoun is not used, the pronoun precedes
the verb like other words ; thus we have in the inscriptions of the Persian
period: Urimizda s'arrutav anacn iddminu (iddnu) " Ormazd has given me the
sovereignty" (Beh. 24), mandatttiv anacn inassunu "they will bring me the
tribute" (Naksh-i-Rustam, 9, 10), anacn Akhnrmazdah litstsurd-ni "as for me,
may Ormazd protect me." The last example but one shows that the verb
may have two objects, the remoter object or dative standing after the nearer
object or accusative. The dative of the pronoun, however, is generally
expressed by the suffixed possessive ; e.g., sane crdni dannnti . . . addin-sii
"two strong cities I gave him" (Khors. 52), xxii. birdti iddin-sn "twenty-two
fortresses he gave him " (Khors. 39). In this Assyrian agrees with Hebrew
and Ethiopic, Where the dative is not a pronoun, however, the preposition
ana is employed, as in dnnkn ana nisi iddinn " he has given prosperity to men "
(N.-R. 2), the compound ana eli being used with a pronoun (as anaeli-su idricuh
" they pursued after him " Beh. 16). In the later dialect of the Achsemenian
122 LECTURE VIII.
inscriptions, when Aramaisms had penetrated into the language, the same
preposition ana is used like the Aramaic "7 to denote the accusative, which
then mostly follows the verb. Thus we read {aducii) ana Gumdtav " (I had
killed) Gomates " (Beh. 109), sa ana khis'Carsah s'arra ibnu "who has made
Xerxes king."' This use of ana is never met with in the inscriptions of
Nineveh and Babylonia, and characterises the Achsemenian period.
The object of a sentence may be implicitly contained in the predicate or
verb, and will then be expressed by a noun formed from the same root as the
verb, and possessing a similar signification. Assyrian is particularly fond of
this construction with "an accusative of cognate meaning," as the gram-
marians term it, and numberless examples might be collected from the
inscriptions. Thus we meet with such phrases as acul acalii " I ate food
continuously," dicta-sun adiic "their slayables (soldiers) I slew" {W. A. I. II.,
67, 9), khirit-su ak/iri " its ditch I dug," ikhtanabbata khiibut nisi sa Assiir " he is
ever wasting the wasting of the men of Assyria" (Sm. A., 258, 113)'.
The use of the infinitive absolute in Hebrew is not dissimilar ; like the
Assyrian construction, it expresses the ideas of intensity and continuation
which naturally arise from the specification of the exact object of the predicate.
In Hebrew (and Ethiopic) the infinitive absolute stands after the verb when
continuance is denoted, before it when intensity is implied. In Syriac also it
stands before the verb when the idea of intensity is to be marked, whereas
Arabic requires the converse position in such a case. It will be seen that
Assyrian in this respect agrees with Hebrew, and the usage seems to go back
to very ancient date. Continuous action is naturally expressed by setting the
object after the verb, while attention would be drawn to the intensive character
of an action by placing the object of it in the foreground. The Arabic usage
is probably of later growth than that of Hebrew or Assyrian. The Assyrian
"accusative of cognate meaning" is sometimes accompanied by the preposi-
tion ana, like the Hebrew infinitive with '7 ; e.g., batiiH-sun va batuldte-sun ana
sagaltu asgul "their boys and maidens I dishonoured." Still more analogous
to the Hebrew construction arc such expressions as anacu dlicniu altacan ana
sadhari liinsu " I gave orders to write an inscription,"' ana epis ramani-su "to
the working of himself," where sadhari and cpis are infinitives, the latter in the
construct state, the former preserving its verbal force, and so retaining its
' Inscription at Elwend, lines 9, 10. ' Inscription at Yan, line 7.
ASSYRIAN SYNTAX. 123
case-termination. These instances will explain the use of the infinitive with
the negative, as in adi la base "up to the not being" (i.e. "till there were no
more"),' iiia la bana "in the not-doing" (i.e. "while I had leisure"), ana la
casad-i ina mati-sii " for the not getting of me (in order that I might not get) to
his country " (W. A. I. I., lo, 45).
The subject may be understood or implied in the verb, or it may be
expressed by a pronoun. Thus the third person plural may be used im-
personally, as in the curious phrase kharsdnu sakutit epis bithri-smm ikbi'uni-su
"it had been ordered him to make snares in the high woods" (literally " the
high woods (for) the making of their snares they had appointed unto him"),^
or it may express the indeterminate third person, as when we read ana mat
Nizir sa mat Lullu-linipa ikabu-su-ni akdhirib "to the country of Nizir, which
they call the country of Lullu-linipa, I drew near" (W. A. I. I., 20, 34) ; and
in a conditional sentence /// ana ziga yusetstsn''u "or should anyone expose
to harm " (W. A. I. I., 70, 11). The third person singular is frequently used
in the same way, and so we have illica " one came," yusapri "one revealed."
These impersonal singulars often take the place of the passive. A subject is
occasionally supplied in the shape of nisu "man," used in a collective sense,
as, for instance, nisu sa mat'Siikhi ana mat Assuri la illicnni "none of the
Nomad Arabs had gone to Assyria." This employment of nisu as a subject
is parallel with the employment of the third personal pronouns as objects.
They could be employed in the same indefinite way as that in which we
sometimes employ "it," in order to sum up collectively all that has gone
before. Ana bit cili la isarrac-si means "to the store-house he does not
deliver them" {W . A. I. I., 27, 36) ; where the feminine singular si refers to
the columns and other palace-decorations which had been spoken of before.
The masculine singular may sometimes be translated "people," as in nsalvi-s
" I caused the people to approach," or usalic-sic " I caused the people to go ;"
where a variant reading has the plural sunn. So, too, in accordance with the
Semitic idiom which allowed a pronoun to be added pleonastically at the end
of a sentence, we read sallut-su va caniut-su ana er-ya Asiir ubla-su "his spoils
and his treasures to my city Assur I brought it" W. A. I. I., 13, 24) ; imut
takhazi-sunu ecint-su " their materials of war I took it " (Lay. go, 65).
As in the cognate languages, Assyrian verbs which denote such ideas as
' In the Inscription of Sennacherib pubhshed by Grotefend, line 31. * IV. A. I. I., 28, 13.
124 LECTURE VIII.
those o{ filling, giving, finding, and the Hke, may take two accusatives, that is
to say two objects one nearer and the other more remote. Thus " Assur,"
Rimmon-nirari declares, malctU lasanan yuntalln'u katd-su "(with) the kingdom
of Lasanan has filled his hand ;" ' and elsewhere we have dahtii imkhar-sitniiti
" the gift he received them " {W. A. I. I., 41, 28) ; sa itstsuru mubar-su la ibah
" which a bird (for) its crossing finds not" {W. A. I. I., 33, 48, 49) ; sa masac
Ili-biahdi khammahi idrupii *' who had burned the skin of Ili-biahdi with
heat " (W. A. I. I., 36, 25) ; an instance of the common employment of two
accusatives where one expresses an idea cognate with that of the verb. Of
course transitive verbs in Shaphel and Shaphael take two accusatives, while
intransitive verbs may be followed by an accusative of similar meaning, like
illica uriikh muti " he went the path of death." Here the object which is
implicitly contained in the predicate is definitely expressed, so that such
accusatives may be said to resemble the use of the definite article with the
noun. A similar explanation must be given of the so-called accusatives of
direction which may follow a verb of motion ; in illicit ritsut-su "they went to
his help," for instance, the object is that towards which the action travels and
in which it finds repose.
Now that we have considered the various relations in which the subject
and the object may stand to the verb, it is time to see what modifications the
verb itself, or rather the action it represents, may undergo. In the first place
the action may be a compound one, consisting, that is to say, of two ideas.
Instances in which such a compound action is expressed by setting two verbs
side by side without a conjunction, are to be found in Assyrian, as in the
cognate tongues. Thus irdn'u illicu kakkar tsnniini "they descended, they
went to dry ground," means "they went down to dry ground," where the
force of the English adverb is represented by a second verb. In the second
place the action may be modified temporally, the time at which it took place
or the relation in which it stood to the subject or object being regarded
differently. To meet such modifications various tenses are used, and Assyrian,
as we have seen when dealing with the verb, was, for a Semitic language,
peculiarly rich in tenses. It is needless here to repeat what has been already
stated in regard to these tenses ; it must be clear by this time what was the
force and use of the aorist, the present, the future, and the permansive, as
' IK A. I. I., 35, 3, 4.
ASSYRIAN SYNTAX. 1 25
well as of the subjective and objective aorists. It may be noted, however,
that, as in Greek, the aorist was employed as an iterative present in similes
and comparisons, what happens at any time being conceived to have already
taken place on some definite occasion, as in cima Rammanii izgum ** as the
Air-god pours." The present, again, may be used instead of the future,
whenever a future event is regarded as being so certain as to seem actually
present, and the imperative is also found in the same sense. Finally in the
third place the action may be regarded modally, that is to say, according to
the manner in w^hich it is regarded by the speaker as happening or likely to
happen. From this point of view the sentence will be capable of a variety of
modifications. I shall here pass over the moods that have been mentioned in
the account of the conjugation, and confine myself to those different kinds of
sentences which arise from the different ways or modes in which the action
and its representative, the verb, may be looked at.
(i) The affirmative sentence simply states a fact, and as it has been
abundantly illustrated by the examples previously given, I shall pass on to
(2) The negative sentence, formed by the two negative particles la
and nl. The Hebrew distinction between ^^ and ^7^^, the first being used
objectively and the second subjectively, has been lost in Assyrian, and in the
place of it we find another distinction, according to which, while la is employed
preferably with nouns, itl is employed with verbs. In the Achsemenian period
of the language, however, /// came to be used with nouns as well as verbs, and
Dr. Schrader may be right in thinking that til was considered more emphatic
than la. As the negative is really part of the idea expressed by the predicate,
it does not require to be represented by a separate word, and Assyrian,
accordingly, possessed a negative verb ydnii, the Hebrew p^<. The primitive
substantival character of ydnu is illustrated by its use with the pronouns ;
ydnii-a, for instance, literally " my not-being" is " I am not."
(3) Deprecatory sentences are formed by the help of the particle ai, the
Ethiopic 'i, Heb. ^{< ; thus ai ipparcit'u idd-sa " may its defences {dual) not be
broken" (Lay. 42, 53), ai isi naciri mugalliti "may I not have enemies
multiplied" [W. A. I. I., 58, 10, 15). In composition with the indefinite
umma at the beginning of a clause, and with id or la following immediately
before the verb, ai signifies "no one whatsoever," as aiumma ina libbi-simu
asar-su ul yinnassi^i "no one among them touched its site" {W. A. I. I.,
126 LECTURE VIII.
36, 36). Hence as the force of the negative lies in the second particle ul
or la, aiuinma came to have a purely indefinite sense when used by itself, and
we may therefore translate it " any one whatever."
(4) A question is rarely found in the texts, and when it occurs has no
such indicative particle as is possessed by Hebrew. In one of the bilingual
hymns (K. 2861) we read: ina same mannu tsirn "in heaven who (is)
supreme?" ina irtsitiv mannu ts'irii "in earth who (is) supreme?" and at
Naksh-i-Rustam (25): matat annitav acca ikilsah "those lands, how different
are they ? "
(5) An intensive sentence may be indicated by the particle In "verily"
placed before the verb, though we also find it attached to a noun, as anacu In
^arru " I (am) indeed king" {W . A. I. I., 58, 9, 62). This use of In led to its
being employed as a mark of past time, like kad in Arabic, so that In allic
means " I went" (at a particular moment), Ifisardi, for /// usardi, " I caused to
add." Other modes of denoting an intensive sentence have already been
noticed.
It would be needless to say more regarding (6) hortative and (7) pre-
cative sentences than that they are expressed by the imperative and precative
moods. The " Descent of Istar " (5, 10) has preserved for us the interjection
allii "woe!" the Heb. ^S'pX.
The simple sentence sufficed for all the wants of primitive man. But
with the growth of his intelligence and knowledge a combination of two or
more sentences with one another came to be necessary. At first these were
merely set side by side ; then the relation of subordination which one of them
implicitly bore to the other was made explicit by some external sign. Con-
junctive and other particles came into use, the relative pronoun was developed,
and the verbal forms underwent many modifications. The more advanced
a language and the intellectual powers of its speakers, the more complicated
become the relations of sentences to each other. Greek and English are good
examples of the manifold ways in which one idea may be subordinated or
related to another, and these relations represented by phonetic means.
Assyrian, like the other Semitic tongues, was very much behind Greek
or English in this respect. The compound sentence remained to the
last comparatively simple, and the stock of modifying particles was not
large.
ASSYRIAN SYNTAX. 1 27
The first mode adopted for connecting two sentences together was by
employing the conjunction va "and," which in many languages, at all events,
originally signified "further" or "addition." That this must once have been
its meaning in Semitic also seems clear from the peculiar Assyrian habit of
commencing a sentence with the verb and inserting the copulative conjunction
between the verb and its subject, a habit, however, which becomes more
frequent the older the inscriptions are. Thus in the account of the Deluge
the Chaldean Noah says usetsi-va s'luuniata "I sent forth also a dove"
(W. A. I. IV., 50, 38). Here the conjunction must be rendered "also" or
" moreover." This half consciousness of its primitive signification will explain
how the copulative conjunction "and" comes to be used as a separative " or,"
and an adversative "but." As in Hebrew, the conjunction vd or u was
weakened into an enclitic with a short vowel {vci) when employed with verbs,
but whereas the Hebrew enclitic is prefixed to the word following, the
Assyrian conjunction is postfixed to the verb (or verbal idea) that goes before.
In this case it might even be contracted into a simple -a. But the Assyrian
va did not possess the flexibility of the Hebrew \ and the only trace of the use
of waw conseciitivum is to be found in sentences in which a subjective aorist is
followed by a construct aorist or a permansive. Such sentences, however,
are common enough in the inscriptions, and may be termed conditional.
Thus w^e have itsbatnni-va eniuru "when they had taken they sa.v/,^' its batuni-v a
tebiini " when they had taken they came." Dr. Schrader remarks that the use
of waw to express a circumstantial sentence seems unknown to Assyrian,
where the present participle is preferred for the purpose ; e.g., ina er Cugiinacca
ina Pars'ii dsib su ina Elaniti itbci (Beh. 41) " in the city of Cugunacca in Parthia
dwelling, into Elam he came."
A conditional sentence may be denoted by other means than the use of
the enclitic va with the subjective aorist. The particle ci is especially
employed for this object, and it may be used with either the future, the
subjective aorist, or the objective aorist. In the latter case, the objective
aorist becomes a true subjunctive. Thus we find ci ecalu ilabbiru-va inakhu
"when the palace shall grow old and decay," ci takabbu'ii iimma " if thou shalt
say thus" (N. R. 25), where the certainty of the event looked forward to is
intended to be brought into relief. It is remarkable that Assyrian alone of
the Semitic idioms should have developed the conditional sense of ci ; in its
128 LECTURE VIII.
use of c'l as a particle of time it agrees with the cognate tongues. But cl was
not the only particle which could be employed conditionally ; besides ivi "if,"
sa is occasionally met with in the same sense, e.g., sa Id agnCu-su igra-ni " as
I did not make war upon him he made war upon me." The subjunctive
enclitic -;/?, however, could be used by itself to express a condition, without
any conditional particle going before ; yutsu-m ner-ya itsbiit, for instance, must
be rendered " when he came out he took my yoke." Here it will be seen that
the conditional sense is almost absorbed by the temporal ; this is not the case
with the sentence ikhkhara abdhu amattu sa p'l-su yustenna " (whoever) evades
the pledge, the truth of his mouth changes" {W. A. I. I., 27, 86), where all
indication of the condition is omitted.
Next to the simple copulative sentence, with the conditional that has
grown out of it, will come the relative, which plays a considerable part in
Assyrian syntax. The relative was originally a demonstrative, as our own
use of that still shows, and such an expression as "he is the man whom I
saw" would once have run "he is the man; that (man) I saw." The mere
juxtaposition of the two clauses was sufficient to evolve the fact that the
second was subordinate to the first, and in course of time the connecting link
between the two, the demonstrative pronoun, acquired a meaning which
expressed this subordination. In the periphrastic genitive, the Assyrian
relative sa, which primarily denoted that the second noun belonged to a new
sentence dependent upon the first, was crystallised into a mere sign of a case ;
elsewhere, however, it better preserved its integrity.
Instances of the ordinary construction of the relative may be found in
almost every inscription of any length. In agreement, however, with the
cognate languages, Assyrian preferred to make the reference to the antecedent
clearer by attaching to the noun or verb which followed the relative the"
possessive or personal pronoun: thus Yahudii sa asar-su rii'kn "Judah whose
place (is) remote," literally "of which its place (is) remote" (Lay. ^2, 8) ;
sa ina abli-su " upon whose son," literally "whom upon his son " (IF. A, I. I.,
35, 3, 2). This repetition of the pronoun, like the repetition of the negative,
characterises a primitive state of speech when the understanding is less
exercised, and accordingly requires language to be more definite and emphatic.
The demonstrative (that is to say, the relative) is not at first felt to represent
fully enough the idea which has gone before. Indeed the demonstrative was
ASSYRIAN SYNTAX. 1 29
to some extent pleonastic ; there was a time, as we have seen, when the mere
juxtaposition of two sentences was sufficient to express the relation between
them. Traces of this time survived in Assyrian, as in Hebrew poetry (and in
our own language), in which the relative might be omitted; e.g., itti kari ab-i
icziivu "with the castle my father had made" [W. A. I. I., 55, 5, 30), assii
khiiltuv ebusii " on account of the wickedness he had done," asar tallaci " the
place (to which) thou goest" (Sm. A.^ 125, 61), yinnu annitu emiirii "the
day he had seen that (dream)" (Sm. A.^ 74, 19). The use of the indefinite
mannu without 5^ is similar; thus mannu atta sarvu "whatsoever king thou
(mayest be)."
The relative may refer either to the whole of the preceding sentence (or
idea) or to only part of it. It may, for instance, relate solely to a possessive
pronoun suffix ; of this we have an example at Naksh-i-Rustam (26), tsalmand-
diinn aniiiru sa cihs'u attu-a nasiCii " behold the images of those who uphold my
throne." On the other hand it may refer to an unexpressed antecedent, and
so be used substantivally, like our "what;" ana sa ebi'isu is "what I have
done," where the preposition is the sign of the accusative, sa anacii ebiis " all
that I did." This will explain the omission of the antecedent in the phrase
ina sa Gargamis "according to the (maneh) of Carchemish," which would be
literally "according to that which (is) of Carchemish" {W. A. I. III., 47,
This quasi-substantival employment of the relative led to its absolute use
at the beginning of a clause, where it summed up all that had gone before.
Thus we have sa ana natsir citte va inisavi-su .... inambu-inni Hi rabi " as
regards which, for the protection of its institutions and laws .... the great
gods proclaim me" (W . A. I. I., 36, 40). Hence comes its adverbial use, as
when we read sa . . . ina cuddi sarruti rabis iisibu "when on my royal throne
mightily I had sat" {W . A. I. I., 18, 44), where the temporal sense of the
passage would ordinarily have required ci. It is remarkable that whereas ci is
rarely used in this way in the other Semitic languages, Ethiopic exclusively,
and Hebrew usually, apply it to the expression of statements of fact which are
uniformly introduced by sa in Assyrian. Thus we get sa Pars'ai ni'tikii ultn
mati-su tsaltav itebus " (when wilt thou recognize) that the Persian far from his
country made war" (N. R. 29), sa la Barziya anacu " that I (am) not Bardes"
(Beh. 21). Dr. Schrader observes that after verbs which relate to the senses
10
130
LECTURE VIII.
a Statement of fact is represented in Assyrian by a nomen inutati or infinitive ;
thus it is said of Merodach-Baladan that halac girri-ya isme "he heard of the
marching of my expedition," where in any other of the Semitic idioms we
should have had a particle such as ^Z*^, or the like.
Sentences denoting the purpose or object may be connected with the
principal clause by a combination of the relative with libbii " heart," " middle,"
or even by means of ina libbi without the relative construction, which Dr.
Schrader has aptly compared with the use of the Hebrew [V^?. We find, for
instance, ina libbi tuma/i-sunutav " in order that thou learn them " (N. R. 27),
libbu sa Gnmdtav agasft Magu-sii bita attu-nu la issiCu "in order that that Gomates
the Magian might not destroy our house." Anama sa is employed in the
same sense, as at Behistun (21) anama sa la yiiniaddiinu " in order that they
might not learn." These, however, were all expedients that belonged to the
Achsemenian age of Assyria; in the older inscriptions such clauses as those we
are now considering were expressed by the preposition ana and an infinitive,
in full analogy wath Hebrew, Aramaic, and Ethiopic. This construction
is also found on the Persian monuments ; ana ebis linisu, for instance, being
" in order to make a tablet." Sa and assu are also found sometimes in the
place of ana ; thus sa limnu la bane panini means " in order that the evil-doer
may not make head," assum aibi la bane panini " in order that the wicked may
not make head."
Two kinds of subordinate sentences are now alone left for our notice.
A sentence may be temporal, denoting the time at which the event recorded
in the principal sentence took place. We have already seen that where
one event followed as a consequence upon another the Assyrians adopted a
construction that reminds us very forcibly of the Hebrew waw consecutivum.
We have seen also that sa might be occasionally used in a temporal sense.
But the chief particle that served for this purpose was ci ; as at Naksh-i-
Rustam (20), ci iniurn "when he had seen." It was through its temporal
sense that ci came to have a conditional one, as in the statement ci tagabbu
" if thou sayest " (N. R. 25). The sense of " until " was expressed by adi with
the future, as in adi allacu "until I shall go" (Sm. A., 125, 67); and in
the Achaemenian period by the compound adi ^eli sa, which, however, had a
more conditional meaning than the simple adi. Thus at Behistun (47) we
have adi ^eli sa anacu allacu ana mat Madai "until I should go (or, "have gone")
ASSYRIAN SYNTAX. I3I
to the country of the Medes." With the construct aorist the compound con-
junction had the sense of "whilst;" e.g., adi ^eli sa agd ebus "whilst I did
this" (N. R. 32). We may compare the Hebrew *ll^*^? "ly or '7 "ly. The
other kind of sentence to which I refer is one which may be termed corre-
lative, an example of which is quoted by Prof. Schrader from Behistun (104),
in the shape of nl anacu id zirya " neither I nor my seed." An affirmative
sentence of the kind is formed, as in Hebrew, by the repetition of the conjunc-
tion va.
Perhaps we should not pass over a very common form of sentence in
Assyrian, in which a quotation is introduced by iimma " thus," answering to
the Greek otl, the Aramaic "*'!, the Arabic ^an, and the Ethiopic cama or ysma.
Instances will be : igabbi uinma anacu ^ar Elami " he says that ' I (am) king
of Elam ' " (Beh. 30), {Istar) ciham ikbi-sunut iimma anacu allac ina makhar
Assur-ban.i-pal " (Istar) thus said to them that * I am going in the presence of
Assur-bani-pal' " (Sm. A., 221, 2, 4), ciham ikbuni unima temenna suati
nubahi la niniiir " thus they had said that ' this foundation-stone we sought, we
could not see' " (W. A. I. I., 6g, 2, 55). The frequent combination of the
two particles, cihaiji at the beginning of the sentence and uninia at its end,
should be observed. Ciham is very common in the inscriptions of the
Achsemenian period, whereas its employment in the older texts is rare.
Before concluding this lecture I would call your attention to the use of
the Assyrian passive participle, to express the sense of " able to be " or
"ought to be," as afia siri sarri "a gate ought to be begun" (W . A. I. HI.,
53, 2). The Pael participles of concave verbs more especially bear this
meaning : thus dicu is " what can be slain," la niba " what cannot be counted,"
pu'u 'ussuru " a mouth that should be bound." Perhaps, too, a few words
might be said about the use of some of the prepositions ; the ideas of
"change," "result," or "object," for instance, are denoted by ana v^'ith the
accusative, as in ana tulle n simmi itur "to mounds and ruins it turned," ana
suzitb napsati-sun ipparsidii " to save their lives they fled." Ina, again, has the
signification of "into" after a verb which means "to descend," and ultu is
occasionally used adverbially for "after that," "from the time when," with
ytimi sa "the day whereon " understood. Thus in the "Descent of Istar"
{Rev. 16) we read ultu libba-sa im'tkhu "after that her heart had rested,"' isUi
' Or rather, " she had rested in her heart," the verb being in the mascuhne instead of the feminine.
10*
132
LECTURE VIII.
ibnd-ni Mariidiic " from the time when Merodach created me." Ultu may also
signify "exacting punishment from," as in ////// Assiiri tirra dude abi "from
Assyria bring back the slaughter of (thy) father," that is to say, "avenge
upon Assyria thy father's death."
What has been said will make it plain how closely the syntax of the
Assyrian language agrees with that of the other Semitic tongues. In some
respects, indeed, it is even simpler and more primitive, and in its temporal
use of ci approaches the idiom of our modern languages. Thought, as
expressed in the sentence, must be either predicative or objective, and the
simplest form of the sentence must be much the same in all languages. The
connection of ideas when reduced to its most necessary elements admits of
but little variation. It is only when the sentence becomes developed and
complicated, and more especially when two sentences are brought into relation
one with another, that syntactical differences on any large scale become
possible. But the syntax even of the simplest sentence is not necessarily the
same in all families of speech. Subject and predicate, or subject verb and
object, admit of varying arrangement, and while some languages (like the
Polynesian) do not possess a verb at all, others (like the Semitic dialects)
possess only what is on its way towards becoming a verb. Even mere predi-
cation may be conceived of differently by different races of men, and accord-
ingly the original Aryan conception which placed the predicate before the
word it defined (as "good man," "man's good") is reversed in the Semitic
languages, where adjective and genitive must follow the subject. Since the
predicative sentence easily passes into the objective sentence, " man is
mortal," for instance, being the same as " man possesses mortality," we find
that the relative position of the subject and object is determined by that of
the subject and predicate. In the Aryan languages the governed word was
primarily placed before the word that governed it, just as the predicate was
before its subject. Similarly in the Semitic languages we should expect to
find the objective sentence following the rule of the predicative sentence, and
making the object succeed the verb. I need only point to Hebrew to prove
how familiar this order was to the Semitic mind, and numberless examples of
it occur in Assyrian. In the latter language, however, it must be 'confessed
that the reverse arrangement had become predominant, the verb being
relegated to the end of the sentence. In this we may, perhaps, see the effects
ASSYRIAN SYNTAX.
^33
of Accadian influence, the Accadian verb regularly standing after its case.
Should this suggestion be correct, we shall have the evidence of comparative
syntax also for the fact which is borne out by the accidence and the lexicon, —
the influence, namely, exerted by the old agglutinative language of Chaldea
upon the Semitic dialects which superseded it.
134
LECTURE IX.
Tlie Affinities of the Assyriati Language and the Origin of Semitic Culture.
OW that we have finished our review of the Assyrian syllabary and
grammar, we can look about us, and consider the conclusions to which
we have been led. Assyrian, we have seen, is a Semitic language which
made use of a foreign mode of writing, and, like the Japanese, which has
similarly borrowed the Chinese characters, had to adapt it to the expression of
its sounds as best it could. The Accadian inventors of the syllabary spoke
an agglutinative tongue, and since the characters they employed were
originally nothing more than hieroglyphics or ideographs, they will inform us
what objects and ideas were known at the time they were invented, and
consequently what degree of civilisation had already been reached. The
Semitic population which succeeded the people of Accad was inferior in
culture, and accordingly borrowed largely from the old race. Not only the
system of writing, but the Accadian literature, along with the elements of
Accadian art and science, became the property of the new comers. So
extensive a borrowing necessarily left its marks on the language of the
borrowers, and we shall therefore expect to find in Assyrian numerous words
which were taken from the alien speech. But when two languages exist side
by side for any length of time, the influence of that spoken by the more
civilised race is likely to extend beyond the mere vocabulary; phonolog>% idiom
and even grammar are all apt to be affected. I have already had occasion to
point out that this seems to have been the case with Assyrian ; the inter-
change of m and v, so characteristic of Accadian, is to be found in Assyrian
AFFINITIES OF ASSYRIAN. 135
also, and the appearance of something Hke real tenses in the Assyrian verb
is best explained by Accadian influence. Of course, the borrowing was not
all on one side ; during the long period when the two races dwelt close
together the Accadians borrowed many words, such as bandar (Ass. passuru)
"a dish," kharub "a locust," s'uccal(}) "a messenger," isib "a settlement,"
adama "the red race," from their neighbours, and Semitic influence will best
account for the fact that the Accadian verb which originally postfixed the
pronouns came afterwards to prefix them, while the adjective followed its
substantive instead of preceding it as was once the case.'
The Semitic inhabitants of Babylonia are called Casdim in the Old
Testament, a word which I would connect with the common Assyrian casadu
" to possess " or " conquer." The Casdim would, accordingly, be the Assyrian
cdsidi or " conquerors," who first made their appearance in Sumir or Shinar,
that is to say, north-western Chaldea, at some unknown period before the
second millenium b.c' The language of these Casdim was what is termed
Assyrian, though the Assyrians were merely a part of the nation which
migrated northward about 1800 B.C., and the Assyrian language merely a
dialect of the Babylonian from which it differed but slightly. A comparison
of the Assyrian and Accadian lexicons, by disclosing the debt of the former
to the latter, ought to indicate in some measure the amount of civilisation
possessed by the Semitic Babylonians when they first came into contact with
the Accadian race.
Here we are confronted by the question : what is the relation of the
Semitic dialect of Babylonia, whether we call it Babylonian, or Assyrian, or
anything else, to the other dialects of the Semitic family of speech ? Is it
most closely allied to the northern Semitic tongues, Hebrew, Phoenician and
Aramaic, or to the southern, Arabic, Himyaritic, and Ethiopic ? The
question can only be answered by an appeal to the grammar and dictionary,
but more especially to the grammar.
From time to time I have had to draw your attention to the analogies
' Many of the words given in the syllabaries as borrowed from the Semites do not appear in the Accadian
texts, and must have been confined to the literary class of a later day, which was partly Accadian, partly
Semitic. Thus I'bila from the Assyrian 'ablie, was employed in the place of the native dunui {dit) or tiir, " a
son ;" perhaps, too, libis " a heart," from the Assyrian libbi-su " his heart," instead of the native siiii or sa.
■ The position of Sumir or Shinar is fixed by Gen. x. lo, which places in it Babylon and Erech, the great
cities of north-western Chaldea. Accad, on the other hand, is sometimes called Uru or Uri ( W. A. I. III., 70,
154), from its capital Ur (now Mugheir) in the south on the western bank of the Euphrates.
136 LECTURE IX.
existing between the grammar of Assyrian and those of the other Semitic
idioms, but particularly that of Hebrew. And, in fact, it is to Hebrew that
Assyrian is most akin. In the first place the phonology of the two languages
agrees in a remarkable manner. The sibilants are not changed into dentals
as in Aramaic and Arabic, and though we sometimes find ^ and D changing
places in Hebrew and Assyrian, this is not the case with the majority of roots.
In Assyrian itself, moreover, certain words are found now with D, now with ^,
while we all remember that the northern Israelites were distinguished from
their southern brethren by their preference for the sound oi samech (Jud. xii. 6).
Then in the second place, there is a striking agreement between the gram-
matical forms of Hebrew and Assyrian. The Niphal conjugation characterises
both, and though the use of the secondary conjugations with inserted t has
attained larger proportions in Assyrian, the Hebrew Hithpael proves that the
starting-point of the two languages must have been the same, Assyrian merely
developing what Hebrew restricted in use. Indeed, in our comparison of
forms we must never forget that the Hebrew of the Old Testament is a very
attenuated speech. It has lost forms which it must once have possessed and
has undergone to a large extent the action of phonetic decay. Traces of the
case-endings are still to be found in it. The accusative of direction in H— still
retains the long vowel which has been shortened in Assyrian, and the nomina-
tive in 1— and genitive in "^— are still to be met with in scattered passages, such
as Gen. i. 24 ; Num. xxiv. 3, 15 ; Psa. cxiv. 8 ; Isa. i. 21 ; as well as in proper
names like Bethtc-el and Penu-el. It is probable that the loss of the case-
endings is often to be ascribed to the alphabet in which the Old Testament is
written, final vowels not being expressed if they were short. At all events the
inscription of Shishak, in which the local names of Palestine, like Negebu
" Negeb," are made to terminate in a vowel, contains a clear proof that the
case-endings were preserved in Hebrew not ver}^ long before the Moabite
inscription of King Mesha was inscribed. In compounds like Penuel their
presence was naturally marked in the writing, and the Hebrew accentuation
which now falls for the most part on the last syllable, though analogy would
require it to fall on the penultima, is further evidence of their former existence
in the language. Even the mimmation may perhaps be detected in such
adjectives as D^"^^ or D^^5S{, if we assume that they are old accusatives, and the
feminine ending in t which has been weakened into H in so many instances
AFFINITIES OF ASSYRIAN. 137
constantly reappears, as in the construct state where it has maintained itself,
like the nominative-ending in Penuel, on account of the pronunciation which
made the construct and the following genitive form together a single com-
pound. Even Assyrian offers a few examples in which the final dental of the
feminine has been dropped, or rather softened into a simple aspirate, from
which we may draw the inference that the dental itself was an aspirated one.
As regards the dual and plural, Assyrian has been less conservative than
Hebrew ; m has been changed to n, the common masculine plural has
altogether lost the last consonant like the construct plural in Hebrew, and the
termination of the dual has been similarly reduced to a vowel and restricted
in use. In the third person plural of the verb, too, the construct aorist
has dropped its final consonant {-un, -an), just as Hebrew has done, though
the subjective aorist retains the older ending {-fini, -diii), which also makes its
appearance sporadically in Hebrew. Even the Shaphel is shown to have
been once common in Hebrew by the crystallised forms ]3ti^ from ]1D, ^pti/
from ^p, etc. ; elsewhere the initial sibilant has become an aspirate, as in the
pronoun of the third person. The Aphel conjugation proves that the same
process had begun to be at work in Assyrian also. In other respects, however,
the Assyrian verb presents resemblances to the Arabic. The various forms
of the aorist, the existence of a precative, of passives in u (like the
Hebrew Pual), and of conjugations analogous to the 8th, 14th and 15th of
Arabic, all remind us of the latter language. But these resemblances
resolve themselves partly into the preservation of primitive forms like
those of the subjective and objective aorists, partly into a similar but
independent development as in the case of the artificial regularity of the
conjugations.
To a later development must be assigned the Hebrew article and
inseparable prepositions. No traces of an article make their appearance in
Assyrian before the Achaemenian period, and the Hebrew article was probably
at the outset nothing more than the demonstrative pronoun, which answered
to the Assyrian 'ullit. In the Assyrian inscriptions we already meet with the
abbreviated li instead of liviti (from m'?), and it instead of itti, but except in
the case of lapaiii (Heb. "'Jd'?), an inseparable preposition cannot be said to
exist.- The analogy of the construct feminine singular and plural and
masculine plural, however, in which Hebrew and Assyrian agree exactly with
138 LECTURE IX.
one another, might have led us to assume a parallel agreement in the case of
the masculine construct singular ; but instances like Penu-el, which would be
pan-il{i) in Assyrian, show that this assumption would have been incorrect.
The Hebrew usage in which the vowel-endings are preserved must have been
the original one, since in Arabic it is only the nunnation and the final -;// of
the dual and -na of the pliivalis samis which are lost before the genitive. It is
probable, therefore, that the shortened pronunciation which caused the loss of
the terminations of the first noun acted primarily on the dual and plural, and
was afterwards extended by analogy to the singular. So far as the feminine
singular was concerned, the loss must have occurred before the Assyrians and
Hebrews parted company with one another ; the masculine construct singular,
however, would have preserved its case-endings, though not its mimmation,
until a subsequent period.
If now we turn to the lexicon, we shall find the most striking agreement
between Assyrian on the one side and Hebrew on the other. Such an agree-
ment will be looked for in vain between Assyrian and any other Semitic
language, with the exception, of course, of Phoenician, which is practically
identical with Hebrew. But even where Phoenician and Hebrew differ in their
use of words, we find Assyrian agreeing rather with the latter than with the
former ; thus " foot " is raglii, '7^"], and not D^/ii ; " good " is dhabu, 21C0, not
DI/3, and the root p^ " to establish," has not passed into the general idea of
"existence," as in Phoenician, ^thiopic, and Arabic. Not unfrequently, how-
ever, words that are archaic or poetical in Hebrew are common both in
Phoenician and Assyrian ; this is the case with alpu, (^bi^) ** an ox," instead
of IM:;; arkhiL, ''a. month,'' (m^) instead of tr'in ; ov pilu, "worked," (^^I/D)
instead of Hi^^. So, too, Phoenician coincides with Assyrian in its use of the
participle as a tense, as well as of the relative t2/ (sa) instead of 1^)^. This
relative, however, was also employed in the northern dialect of Hebrew, as
may be seen from the books of Judges and Canticles.
Aramaic, the remaining member of the North Semitic group, stands at
a great distance from Assyrian. Indeed, it differs from Assyrian in almost
all those points in which it differs from Hebrew. Its phonology has undergone
a considerable change, a good many of the sibilants having become dentals,
while tsaddi has sometimes passed into ^. In the Aramaic p"^}! and y^\T^ it
is difficult to recognise the Assyrian irtsi(tuv), and sanu'u, or the Hebrew rii^
AFFINITIES OF ASSYRIAN. 139
and D'^iJ^- Though the consonantal system of Assyrian itself has undergone
considerable modification, as we have seen in an earlier lecture, kheth being
frequently omitted and s changing into / before a dental, in all the points in
which Aramaic has departed from the primitive Semitic phonology, Assyrian
remains true to the earlier sounds. On the other hand it must be allowed
that the latter language resolves a final H into u, just as Aramaic does into i^,
while the guttural sound of ;/, so characteristic of Hebrew, is almost, if not
entirely, unknown to Assyrian. But the " emphatic aleph " or postfixed
article of the Aramaic idioms, has nothing analogous in the language of
Nineveh, and implies a previous loss of the case-endings. The preservation
of a Shaphel conjugation is one of the marks of archaism which Aramaic
shares with Assyrian ; the formation of the precative, being common to Arabic
and Ethiopic, as well as to Aramaic and Assyrian, would also go back to the
period that preceded the separation of the Semitic tribes ; and the peri-
phrastic genitive is found in all the Semitic tongues. The loss of the emphatic
aleph in the construct state is easily explicable from its origin, and is not to
be compared with the loss of the Assyrian case-endings in the same position,
while the Assyrian use of ana with the accusative, and the mode in which the
superlative is denoted, belong to the later period of the language when it had
been affected by Aramaic influences. The employment of di, the passives in eth,
the want of a Niphal, the dropping of the vowels, the extension given to the
formation of abstracts, the use of compound tenses, and of the substantive
verb nii^ instead of ::^i (isn), all draw a clear line of demarcation between the
idiom of Syria and that of Assyria. The vocabulary, too, points in the same
direction. " Man," in Assyrian, is adimi, (D^^5) rather than mi)^ ; "to take," is
np'?, rather than hyp\ " king," is "j'?^, rather than dl^ ; while the specifically
Aramaic -)1, "son," is replaced by 'ablu, (py) and binu (]n).' Aramaic must
have separated from its sister-dialects and entered upon an independent
course of development long before the ancestors of the Hebrews and the
Phoenicians had quitted their kinsfolk in Babylonia. And this is borne out by
tradition. The Phoenicians believed that they had originally migrated from
' At the same time, as might have been expected from their proximity, the vocabularies of Aram and
Assyria contain a considerable number of words in common. Thus we have the Assyrian elipptt {elipu), "a ship,"
Aram. XD^X ; wrt'/?/, " country " (of Accadian derivation), Aram. SnD ; /^rt'rw, " heap," Aram. Kir ; and
Dr. Delitzsch notices that the Assyrian talinm is the Samaritan tellem, " full brother," and the Aram, telaina,
which we have also in the proper name Bartholomew (" son of Talmai"). See, also, Num. xiii. 22.
140 LECTURE IX.
the Persian Gulf,' Kepheus ruled in Chaldea, according to one legend, and at
Joppa according to another, and the Israelites never forgot that their father
Abram had been born at " Ur of the Chaldees." But we look in vain for any
traces of a similar tradition amongst the mountainous tribes of Aram, or
'Subarti, the " highlands " as it was called in Accadian. There was, it is true
an early connection between Babylonia and Kharran, which is itself an
Accadian name, meaning, "the road;" Dun-ciin-tiddit , or Mercury, is termed
"the spirit of the men of Kharran" {W. A. I. III., 67, 28); and Sargon
declares that he had restored "the decrees of Assyria and Kharran, which from
distant days had been set aside, and their laws neglected " (Botta, 144, 11) ;
but we need not look beyond the statement in Gen. xxxi. 47, to see that an
essential difference was felt to exist between Aram and Canaan. It has
long been recognised that the table in Gen. x. is geographical rather than
ethnological, arranging the nations of western Asia according to their
position, not according to their descent.
At the same time, Aramaic belongs rather to the northern branch of the
Semitic family than to the southern, which comprises the Arabian of central
Arabia, and the Himyaritic or Sabean of Yemen, along with the Gheez or
Ethiopic of Abyssinia. The characteristic feature of the southern group is
the existence of broken plurals, originally collective singulars, which are
altogether wanting in the northern section of the family. The vocabulary,
again, marks the southern branch off from the northern, and we may point to
the name of the numeral "six," which retains its medial dental in Arabic and
Ethiopic (shadash), as a further evidence of the same fact. The consonantal
system of the southern group, moreover, differs from that of the northern in
having developed new sounds. Arabic, however, has been singularly con-
servative in regard to its nominal and verbal forms : the mimmation has
become a nunnation, though preserved in one of the dialects of the Himyaritic
inscriptions, and the three case-endings may still be heard, it is said, from the
lips of the Bedouin. The modified forms of the imperfect or aorist, the
passives in -?/, the use of the participle, the adverbs in -c7, the dual in the verb,
the secondary conjugations in / and tan^ and the simplicity of the vowels, are
all so many archaisms which Arabic shares with Assyrian. In Assyrian they
■ vStrabo, i. 2, 35 ; xvi. 3, 4 ; 4, 27 ; Justin, xviii. 3, 2 : Pliny. H. N., iv. 36 ; Herodotus, i. i ; vii. 89 ;
Schol. to Horn., Ocf., iv. 84.
AFFINITIES OF ASSYRIAN. I4I
were stereotyped by becoming part of a literary language; in Arabic they have
been preserved by the nature of the country and the unmixed blood of the
speakers. But even Arabic has not been free from the action of phonetic
decay and other causes of change. To say nothing of the broken plurals or
the contracted forms of the third person plural of the verb, an article has
grown up as in Hebrew ; auxiliary tenses have come into use ; the accent has
been uniformly thrown back as in the Latin language, or the ^olic dialect of
Greece ; and the cases have fallen away in the dual and plural, -dni and -aini
being alone left in the dual, and -iina and -ina in the plural.
If we now turn to Ethiopic we shall find several points in which it agrees
remarkably with Assyrian, while at the same time preserving its character as a
member of the southern group of Semitic tongues. In the first place, the
imperfect has been differentiated into two tenses, one yenger, the Assyrian
isciin, and the other yendger, the Assyrian iscicin. In the second place, the first
person singular of the perfect is formed, as in Assyrian, by the guttural
{gabarcu), though in the second person where Ethiopic has again a guttural,
Assyrian has the dental of the other northern dialects. Then thirdly, the tens
in both Ethiopic and Assyrian are characterised by the same suffix a (e.g.,esrd
Ass. isn't " 20," s'alasd, Ass. silasd " 30 "), and Dr. Schrader notices that it is
in Ethiopic and Assyrian alone, that the old plural ending in -an is shortened
to -d when a noun is used with a numeral denoting one of the tens. Add to
this the existence of an Istaphal, of adverbs in -a, of a suffix -tu or -//, and of
verbal nouns like mafrey (corresponding with the Assyrian manzazii), and we
have a series of remarkable resemblances between the two languages. The
violent letter-change and peculiar prepositions, too, which distinguish Ethiopic,
are analogous to what we meet with in Assyrian. These resemblances, how-
ever, may all be explained as resulting either from the preservation of old
forms which must once have been possessed by all the Semitic idioms, or from
the action of similar circumstances, Ethiopic, like Assyrian, being an offlying
branch of the Semitic stock brought into close contact with an unallied
language. The two forms iscim and isdcin must have been a common heritage
of the Semitic family, while the first personal pronoun an-acu shows that the
form gabarcu or sacnacu is at least as old as the form kabaltu or kdbalti. The
Ethiopic, or Sabean section, would have been separated from the parent
speech while the perfect or permansive was still in the process of making, and
142 LECTURE IX.
for reasons which it is impossible to discover, Assyrian alone of all the dialects
which were left behind continued to prefer the formation with the guttural to
that with the dental. As for the mimmation and the retention of the initial
sibilant in the third personal pronoun, which characterise one of the Him-
yaritic dialects, they are simply survivals from the primitive past.
This brief sketch of the relations of Assyrian to the cognate languages
will have abundantly illustrated its importance for the study of comparative
Semitic grammar, and the light thrown by it upon the parent Semitic speech.
Thus all doubt has been removed in regard to the original existence of the
case-endings in all the Semitic dialects, and I have already endeavoured to
trace the genesis of the tenses of the verb by the help of Assyrian, while I
hope hereafter to show by the same means, that the accentuation of Ethiopic
approaches more nearly that of the parent-speech than does the accentuation
of any other Semitic tongue, Assyrian alone excepted. We are taken back to
a time when as yet there was no verb, or rather no distinction between noun
and verb, when the relative and the periphrastic genitive did not exist, when
the noun was provided with a mimmation as well as a vocalic case-ending
which was not yet dropped in the construct state, when the plural terminated
in -dniu^ used alike of masculine and feminine nouns, and when the accent
fell, for the most part, on the final or penultimate syllables of the word. When
once the reconstruction of primitive Semitic grammar has been made fairly
complete, we may proceed to compare it with Old Egyptian or the sub-Semitic
dialects of northern Africa, and determine how far the resemblances that
seem to exist between Semitic and African grammar are illusory or founded
on fact.
But important as Assyrian is for comparative grammar, it is equally
important for a reconstruction of the primitive Semitic dictionary, and thereby
of primitive Semitic culture. If once we know by the help of comparison
what words were possessed by the Semites before their separation, we shall
have a clue to the degree of civilisation they had reached. We cannot, it is
true, infer from the absence of any words in the later dialects that they had
never been possessed by the parent-speech ; what we can infer is, that where
such words can be proved to exist, the objects or ideas they represent must
have been known. And the Assyrian inscriptions take us back to a time
when Semitic civilisation was growing up under the fostering influences of
AFFINITIES OF ASSYRIAN. I43
Accad. The records of the teachers of the Semitic race have come down to
us, written in their own language and embodying their own thoughts. And
when we have proof that a Semitic word has been borrowed from the
Accadian, we are justified in behoving that the object or idea which it
signified was borrowed also. Dr. Schrader has long ago noticed that the
parallelism which characterises the poetry of the northern Semites has its
prototype in the poetry of Accad, and the epical literature of Babylonia and
Assyria certainly had its source in that of the Accadians. The debts incurred
by the Semitic lexicon and culture may be illustrated by a few examples.
We should expect, however, to find the debt largest in the case of the
Assyrians, and next to them in that of the northern Semites, the southern
Semites of Arabia being least affetted by the Turanian civilisation of the
Euphrates valley ; and such turns out to be the fact.
The city is the first requisite of settled and civilised life. The walled
TToAty with its temples, its marketplace, and its baths, was to the Greeks the
sign and symbol of an organised state. But the Semite, uninfluenced by
favouring circumstances, has ever been a nomade and a wanderer. The
Bedouin of Arabia is the purest specimen of Semitic blood with whom we
can meet, Hebrew tradition brings the patriarchs before us as roving shepherds,
and the " wandering Jew " is still a representative of the best part of the
Semitic race. A pastoral life and trade, these have been the two passions of
the Semites from the earliest times. It is instructive to compare the history
of the mixed population of Babylonia, which preferred to live quietly at home
occupied with agriculture and learning, with that of the purer Assyrian, whose
armies overran the larger portion of western Asia with little other object than
the mere desire of traversing the earth. Now the word for "city" (l^I/)
which is found in Hebrew and Aramaic, and probably forms part of the name
of Jeru-salem, is not met with in Assyrian except as a proper name. This is
Urn, now represented by the mounds of Mugheir, the Ur of Genesis where
Abraham was born. But Uru was of Accadian origin. It was, in fact, the
capital of Accad, or south-eastern Babylonia, and it obeyed the rule of
Accadian princes long after Sumir or Shinar had fallen into Semitic hands.
Uru meant simply " the city," and under another form, that of eri, or rather or/,'
' The original form of cri is given as erini, and translated by the Ass\rian isittu, "foimdation."' Erivi
was adopted by the Assyrians under the form of erinundtu, and with the special sense of "boundary-stone."
144 LECTURE IX.
must have been borrowed by the ancestors of the Hebrews and Aramaeans
before their migration to the west. Like many other Accadian words which
originally began with the syllable ;;;//, Urii has lost its initial consonant,
murit becoming first wiirii and then 'iini. The definite case miiruh, how-
ever, formed by the demonstrative pronoun bi, has preserved the original
labial.'
While the Aramaeans and Hebrews took with them not only the Accadian
conception of city life, but also the name the Accadians gave to it, the
Assyrians, or rather the Semitic Babylonians of Shinar, had adapted a word
of their own to the same purpose. This word was illu, the Hebrew '7^^^ "a
tent," the tent of the nomade being changed into the city of the settled
burgher. The southern Semite remained ignorant of both conception and
word, and when he afterwards began to build his towns and to call them by
a special name, it was one which had no connection with the civilisation of
ancient Accad.
But the existence of the city brought with it new conceptions and con-
sequently new names. The shepherd or the trader had all the world before
him ; he recognised neither boundaries nor landed possessions ; the desert
was limitless and the Semite was free to wander where he would. Settled
life, however, brought with it the recognition of property, the limitation of
landed rights, and the demarcation of state and nation. The Assyrian term
for "country," matii, as opposed to irtsitn (r"l^^) ''the earth," was one borrowed
immediately from Accad. The Accadian ma, " land," was extended into mada •
by the individualising affix da; and while ma represented " country " in general,
mada was some one country in particular. Through the general decay that
affected the terminations of Accadian words mada came to be contracted into
mad^ and this when adopted by the Semites was furnished with a feminine
termination, and so became successively madatii, madtii, uiattii, and mdhi.
The Aramaeans carried the word away with them under the form of i^t^f2, from
which we must infer that the borrowing had taken place before the separation
of the northern Semitic tribes. Pre-Aryan Media, the cradle of the Accadian
race, probably received its name from this word mada, and the "Median"
dynasty of Berosus, which has formed the basis of so many historical and
■ IV. A. I. II., 30, 17. It is possible, however, that murub here denotes ''the woman" (Assyrian 'urit,
mn), regarded as " the conceiver."
AFFINITIES OF ASSYRIAN. I45
ethnological theories, may really have owed its title to the " land " of Shinar
or Accad from which it came.
Nanga ** a district," is another word which made its way from Accadian to
Assyrian, where it took the form of nagii. Since Accadian ng between two
vowels presupposes an earlier in, we may restore the original form of the word
as nama. The word belonged also to the dialect of southern Elam if we are
to judge from the names of the cities Nagitu and Nagiti-dihiban, at the
mouth of the Eulaeus, to which Merodach-Baladan fled for refuge.
The terms applied by the Semitic Babylonians to the foundations of their
brick buildings were naturally borrowed from their Accadian instructors.
Thus epinu "foundation," is the Accadian dpin (contracted into pin), and
temennu "a foundation-stone," is the Accadian timmena (i.e., timena) which
was successively weakened into timmen, timme, tim or tern, and te. It is curious
that, while the Accadians called their Semitic invaders by a name of native
origin, lugud "the white race," from liiga or lugur "man" (possibly for
mtUtcga), and ltd or nda " white," they adopted from their enemies the name of
adamatu "the red race," by which they were themselves called, under the
form of adama. The Accadian language showed the same dislike to the
pronunciation of a final consonant as does modern French, and adama stands
for adamat. We can hardly refuse to accept the old opinion which connected
D^^^ " man," the Assyrian admit, with the root which means to be " red :" in
this case the Adam of scripture would appear to have been Accadian just as
much as " the sons of Elohim " (Gen. vi. 2,) to have belonged to the " white "
Semitic race.
Ippu or ibbn, the Hebrew n^\ was, however, the usual Assyrian word
for " white," and it answered to the Accadian uknu " crystal-white." Just as
the feminines adamatu or tenisetu (ti'liJ^), are used to express the abstract
conceptions of "red race" and " mankind," so the feminine ippatu would be
employed in the sense of " the white race." Now ippatu corresponds letter
for letter with the Biblical riD^ or Japhet, and the question accordingly arises
whether the name of Japhet does not denote that he was the forefather of the
"white" Aryan race. M. Harkavy suggested some time ago that the name
Japhet was to be connected with that of Mount Niphates, and the Aryan root
snig h, hom which we find vlcpa "snow" in Greek, and nix in Latin. The
suggestion is confirmed by the fact that DPI Ham, the father of the swarthy
11
146 LECTURE IX.
Africans, seems to get his name from the root Din (DDn), *^ "to be black
(hot)," the Coptic kham or kam, while Shem must, I think, be connected with
the Assyrian ^amu " brownish." 'Sainii also appears under the form of ^ihauiu,
with which the Hebrew DHJi' must be allied, Assyrian s' answering here to
Hebrew 12/ as in many other instances. Even in Assyrian itself we have iibitti,
"seven," by the side of sibitti; ^arru, by the side oi sarru. 'Samu is applied to
any neutral colour : a blue mist or cloud is called daniu, just as much as a
dark-green stone or a yellow flower. 'Samu is also given as the equivalent of
adru, "dark," from which comes the name of the cloudy winter month Adar, as
well as adirtii, " an eclipse."
Now both damn and adni are, I believe, of Accadian derivation. A
syllabary [W. A. L \l., 1. ijy, 178) tells us that ^]} when sounded dir in
Accadian represented the Assyrian adru, and when sounded ^d the Assyrian
dd'anni ; and elsewhere the same character is rendered by klialapu and siitnini,
" covered," and mikit-isati, " the burning of fire." Just as the original form of
the Accadian pin was dpin, so the original form of dir may have been adir,
from which the Semitic adirii, adru would have been derived ; at all events the
general analogy of Accadian phonology leads us to infer an earlier form, ^am
for s'rt, from which s'a'aniu (and then the weakened s'ihainit) was borrowed." It
is a familiar common-place that semi-barbarous peoples are unable to
distinguish between any but the most obvious colours ; to this day the same
word means both "blue" and "green" in Welsh, and the Homeric 7rop(f)vpeo9,
ohoyJA, and the like, are of the vaguest possible signification. A nice apprecia-
tion of tints shows a fairly advanced state of civilisation. If, then, the
Semites received their first lessons in the art of distinguishing accurately
between colours from the more cultured Accadian, it would only be what we
should expect. And the Semitic name of another colour, yellow, seems
equally referable to an Accadian source. In W.A.I. II., 26, 50-55, ara is
given together with dizi as the Accadian equivalent of the Assyrian arku
"green" or "yellow," and iirik as the equivalent of nrcitu "verdure," while
khir is further translated by the Assyrian arku. A comparison of the three
forms enables us to restore the original ^/z/r/A', which became khir, urik and
' In IV. A. I. II., 26, 47, samanu has been placed in the Accadian column either by mistake, or else
because it had been borrowed by the Accadians from the Semites after the establishment of the latter in
Shinar. In the preceding line ^«^ is given as the Accadian for "blue;" compare the Tatar kuk, ''blue;"
and Protomedic an-cic, " divine blue," or " sky."
AFFINITIES OF ASSYRIAN,
147
ara, the Semites adopting the word while it was still pronounced khirik or urik.
The adoption must certainly go back to a very early time, as the word in
some form or other is to be found in the southern as well as in the northern
Semitic dialects. In Ethiopic, indeed, it has given a name to " gold" {warek).
Another word which was borrowed by the Semites while they still
encamped on the western banks of the Euphrates, before their separation,
was the Accadian ega " a crown." Royalty and its insignia were of Accadian,
not of Semitic growth ; the Semite was by nature a democrat. Mir was used
in the same sense in Accadian, and was also applied to the "halo" seen
surrounding the moon in cloudy weather ; ega, however, was the only term
adopted by the Semitic nomads. From it we have not only the Assyrian agu,
" a crown," but various words signifying round objects, out of which was finally
developed the Arabic verb _U "to be round."
The "great king" himself ruled over Accad before he ruled over Semitic
Shinar, The Hebrew ")ti^ is the Assyrian s'arrii, with this difference, that whereas
in Hebrew "|'7Q is " rex," and "iJi' merely " regulus," the converse is the case in
Assyrian. Now the Assyrian s'arru must, I believe, be referred to an Accadian
origin. 'Srt in that language means "to judge," and ^J^j ^ which we are
informed by glosses is to be pronounced either s'a-galum or da-gar, literally,
"judgment-maker," signifies " monarch." But from sa might also be formed
dara by the sufhx -ra, and we find ^J^f accordingly pronounced in Accadian as
sara, and rendered by the Assyrian sarrii. It does not need to be pointed out
that the derivation of darrii (llJi^), originally a monosyllable and so contrary
to the general character of Semitic roots, is most obviously to be sought in dara.
We shall thus be able to understand how it came about that the Babylonians
and Assyrians who inherited immediately the traditions of Accadian culture,
used darrii in its Accadian sense, while the more distant Hebrews allowed the
native "['^^ to take its place. It must be remembered, however, that a
favourite title of both Accadian and Semitic princes was ri'u, Accadian diba,
" a shepherd " (Hi/l), which reminds us of the Homeric 7roL/j.r)i' XaSu, while
according to Berosus, Alorus, the first of the antediluvian kings, assumed the
title of "shepherd." Such a title certainly suits a pastoral race of nomads
better than the organised communities of pre-Semitic Chaldea.
However this may be, the conceptions connected with the regular adminis-
tration of law might be expected to have emanated from Accad, where it was
ir
148 LECTURE IX.
carried to high perfection, and where, as we have seen, his judicial office gave
the monarch one of his titles. We need not be surprised, therefore, at finding
the semi-monosyllabic jT {]M) "judge," claiming a non-Semitic parentage.
The Accadian di is the equivalent of the Assyrian dinu " a judge " {W.A.I.
II., 7, 32), and di in Accadian presupposes an earlier din and a still earlier
diin{d), like dii {dun, diini) "to go," and other words. The earliest code of
laws of which we know is an Accadian one, and the legal phraseology and
procedure of Accad was very complete.
A good idea of the organised administration of the country may be
obtained from a bilingual tablet given in W. A. I. II., 38, i Rev. Here we
find that " the payer of tribute " {da-ln-u sa bil-ti) had a distinct name, though
unfortunately only the two last characters, ci-ta, of the Accadian word are
left. So, too, had " the defaulter " or khi-bu-u, whose name is written
ideographically in Accadian "the man who makes default," and without any
clue to its pronunciation; and the "tax-payer" or ra-pi-ku, with which the
Aramaic p21 rnay be connected, was similarly described in the Accadian
mode of writing, as "the man who makes payment" (lugur al gara).
After the defaulter comes "the taxgatherer," ma-ci-^u in Assyrian, like the
Arabic ^^.X*, whose Accadian title was, " the man who makes execution "
{lugur gar-tar-da gard) . Next follows "the commissioner of the brickyards,"
la-bi-in la-bit-ti, a very important personage in a country which depended so
largely for writing as well as for building purposes upon its native clay. His
Accadian title we have no means of reading ; it is ideographically written,
LUGUR mur zi-gab, " the man who oversees the bricks." After him we have
" the collector of the taxes," la-kidh kur-ba-an-ni, where it is interesting to
find the corban of the New Testament employed in the sense of " taxation."
The Accadian equivalent is d.p. lak' ririga which bears the same sense. The
alien was termed a-si-bu, " the squatter," in Assyrian, and lugur ca-ca-ma,
" the man told over," in Accadian, while the burgher was ca-tu-u in Assyrian,
and lugur ca gina, "the man of the fixed face," in Accadian. The " tribute"
paid by subject populations was called in Accadian, gun, which is written
phonetically ^^ t^JJ^ gu-un. The word may be allied to the verb gin, or
gen, " to establish."
With all their culture, however, the Accadians were an agricultural rather
• See IV. A./. II., 2, 373.
AFFINITIES OF ASSYRIAN. I4g
than a commercial people, and though we hear of " ships of Ur " {W. A. I.
II., 46, 3), trade does not seem to have been in a very developed state until
after the Semitic conquest. Even the usual word for "price," sam, was
borrowed by the Accadians from their Semitic neighbours. On the other
hand, the old population of Chaldea was famous from an early period for its
mathematical studies, its astronomical observations, its astrology and its
magic, and in the case of all these we should expect to find that the Semites
had borrowed largely, not only ideas, but the words which expressed them as
well. Our expectation will not be disappointed. As I noticed several years
ago in my Assyrian Grammar for Comparative Purposes, the Semitic term for
*' one hundred," Hi^D, Assyrian me, is of Accadian origin. While the
Accadians used a convenient system of cyphers for all numbers under one
hundred, they represented the latter numeral by y»-, me, the ideograph of
"assemblage." Me is a contracted form of eme "a tongue," of which
the ideograph was originally a drawing, and hence it means, " a voice," "to
call," and " an assemblage." But it is also a contracted form of ines, " many,"
which frequently marks the plural in Accadian, like mas in Protomedic, me
itself performing the same function in Susianian. Me further signified "one
hundred," but whether as a later abbreviation of eme or of 7nes I cannot say.
Since H^^D is common to all the Semitic idioms it must be one of those
words which were adopted by the Semites before their separation, when they
had not as yet crossed the Euphrates. This is not the case with another
word, the Assyrian estin, " one," which Dr. Delitzsch has traced with great
probability to an Accadian source.' This is the Accadian as, " one," with the
usual suffix ta-a-an "sum," or "number." Estin makes its appearance in
Hebrew in the name of " eleven," 'ashte dsdr, and we may perhaps infer that
the word was borrowed only by the northern branch of Semites. An example
of the debts incurred by the Semitic dictionary to the Accadian in the matter
of magic will be found in the Hebrew y\)^, primarily a "familiar spirit," and
then, " one who has a familiar spirit." ' The Assyrian equivalent is ubutu, or
abiitu, " magic," which, as M. Lenormant first noticed, comes from the
Accadian ubi " the calling up of a ghost."
It is needless to refer to the evidence borne by the lexicon to all that the
' George Smith's Chaldaische Genesis, pp. 277 sq.
' See Baudissin : Studien zur Semitischen Religionsgeschichte (1876).
I50
LECTURE IX.
Semites owed to their more civilised predecessors in the way of writing and
literature, the ordinary term for " a clay tablet," diippu, or dippu, for instance,
being naturally of Accadian descent ; nor need I weary your patience by
dwelling any longer on the revelation made by cuneiform research as to the
origin of the larger part of Semitic culture. Such a revelation is all the
more unexpected, inasmuch as modern scholars, who mostly belong to the
Aryan or Semitic families, have been in the habit of assuming that a people
who spoke any other than an inflectional language must necessarily be of an
inferior type. It is true that China, Mexico, and Peru, or even the Finn, the
Magyar, and the Turk of the present day, might be cited against such an
assumption ; but race prejudices are always strong, and the facts that bear
against them are never admitted, except after severe opposition and criticism.
Even a philosopher of " common-sense," like Dugald Stewart, once proved to
his own and others' satisfaction that Sanskrit was an artificial language
invented by the Brahmins to deceive the students of the west ; time has
shown, however, that the " unphilosophical " and "deluded" students were
after all right, and that the critic and his friends were wrong.
But it is not only the origin of Semitic culture that has been revealed
by cuneiform research, the nature of the cuneiform system of writing
also enables us to discover the degree of civilisation possessed by the
Accadians themselves at the time of its invention. Since every character
was once a hieroglyphic representing an object or idea, all those objects or
ideas which are expressed by simple (and in some cases by compound)
ideographs, must necessarily have been known to the inventors of the writing.
And further, since the aim of the inventors must have been to give visible
representations of all the objects and ideas with which they were acquainted,
wc may infer that whatever objects or ideas are not so represented must have
been unknown to them. Now an examination of the syllabary will lead to
the following results. The primitive Accadians were polytheistic, but their
worship had already assumed a stellar character quite in accordance with the
other indications that we have of the great antiquity of astronomical observa-
tions among them. "A deity" is symbolised by a star, "a constellation"
by three stars. "The year," too, was already defined, as well as "a month
of 30 days ;" and an incipient knowledge of mathematics is shown by the
existence of ideographs for "number," and "measure." "Law" was
AFFINITIES OF ASSYRIAN. I5I
administered by the "judge," and "the state" was governed by "a monarch,"
who received an annual "revenue" from his "subjects." "Slaves" were
kept, and the existence of "ships" implies also the existence of sailors.
The people dwelt in "houses" of "brick," "wood," and "stone;" these
houses were provided with "doors," "beams," and "seats" of wood, and
possessed " gardens " in front.' A group of houses constituted " a city,"
which had "walls," "gates," and " a citadel ;" the latter, it would seem, was
originally built of wood.* The cities were connected together by "roads,"
which crossed one another, and the country was intersected with " canals."
It is probable, however, that the characters denoting the latter were added to
the syllabary after the Semitic occupation of Shinar, since curiously enough
most of the words denoting them {balag, biting^) are of Semitic origin. The
temple of the deity resembled an ordinary house; but it contained "a
shrine," " an image," and an "altar ;" the royal palace was simply "a large
house." "Carriages" with "wheels" were used, to which oxen were "yoked."
The horse was a subsequent importation from the east, possibly from the
Aryan tribes of the Hindu Kush ; at all events, its Accadian name was
" animal of the east." " Oxen," however, were employed from the first, as
well as the ass, which was emphatically termed "the beast," implying that it
once held the place afterwards occupied by the horse. The other animals
known to the primitive Accadians were "sheep," looked after by " shepherds,"
" the gazelle," "the antelope," "the bear," " the wild bull," " the dove," "the
snake," " the fly," " the flea," " the moth," and some species of " fish."
Bees, too, were plentiful, and their "honey" was an article of food. It is
plain from this list of animals, that the Accadian hieroglyphics were invented
in a mountainous and comparatively cold country.
This agrees with the meaning of the name Acada, or " highlander," which
is formed from the verb aca "to be high," by the individualising suffix da, the
Assyrian equivalent of "the land of the Accadians" (borrowed afterwards by
the Accadians) being tilla, or " the heights," from T^bV•^ As I have pointed out
' That this was the position of the garden or shrubbery is evident from the figure of the character which
represented an " enclosure," or "homestead." Even in its Assyrian form k^^^T the fact is shown plainly.
' At all events, >-^TttT " a fortress " {manzazjt in Assyrian), was called gis-gal, or ''great woodwork,"
in Accadian.
5 ^. A. I. II., 48, 13.
152 LECTURE IX.
on page 43, the earliest writing material seems to have been papyrus
rather than the clay of the Babylonian plain, while M. Oppert has noticed
that the absence of any simple ideograph for the palm shows that the
inventors of the writing must have lived in a colder region than Chaldea.
That this colder region was Elam is made evident not only by the name
"Accadian," but still more by the fact that the same ideograph, 'X'*', denotes
indifferently "a country," and "a mountain." Negative evidence also on
the same side may be found in the fact, that while "a stream" was represented
by a special ideograph, a river was not, an inconceivable occurrence in
Babylonia with its two great rivers. So, too, "bitumen," the peculiar product
of Chaldea, had no representative in the original collection of ideographs. In
fact, the native legends which looked upon the "mountain of the east," the
present Elwend, as the peak whereon the ark rested, and the cradle of the
Accadian race, contained an element of truth.
Besides the animals already mentioned, the inventors of the cuneiform
characters were also acquainted with some kind of cereal, with a sort of
" beer," and with the three metals, "gold," "silver," and "bronze." Silver
was called " the shining," babar (for barbar) ; and the Euphrates in the
neighbourhood of Sippara, was entitled the river of " bronze," [iirudu,
Semitised urudtu). Only meteoric iron was known, whence it was ideo-
graphically denoted by the name of the god Adar. It is noticeable that the
vine appears to have been first met with in Babylonia ; at all events, it was
termed "the tree of life" {ges-tin), and not expressed by a simple ideograph.
Among precious stones " the diamond " was known, as well as its powers of
cutting.
The inventors of the syllabary were armed with " the sword," and " the
bow," which was of course accompanied by " the arrow," and " the quiver ;"
they wore "signet-rings" on their fingers, and "bracelets" on their arms;
and dressed themselves in "linen" or "woollen" robes with "sleeves,"
sometimes dyed "purple," sometimes "variegated;" while their heads were
covered with "turbans." They also used "cups" and "buckets," and
"papyrus" for writing upon; and their sorcerers prepared various kinds of
" poisons." " Witchcraft," indeed, flourished ; the national cult was Sha-
manistic, and it was believed possible to call up the dead. Every object had
its "spirit," and "hymns" were composed in honour of the latter. The
AFFINITIES OF ASSYRIAN. 1 53
country contained " marshes" and "jungle," and "the desert" had received a
name. The mother occupied the chief place in the family, as is evident not
only from an old table of laws given in W. A. I. II., lo, and translated by
myself in the Records of the Past, IV., pp. 21, 22, as well as from the Hymn to
the Seven wicked Spirits,' where the Accadian order, " female they are not,
male they are not," has been reversed by the Semitic scribe, but also from the
ordinary Accadian word for " mother," (C yj^j ) which appears sometimes as
dagal, sometimes as dainalla (i.e., danidia), or damal (H/./4./. IV.,9, 24, 28). Now
both dagal and damal signify " the mistress of the house," being compounds
of dam " mistress," and mal or gal " a house." Mai is connected with mal or
mar "to dwell," and perhaps also mar "a road," (as in martu "the west,"
literally " the path of the setting sun"), gal with the verb which signifies " to
exist." Both mal and gal, originally malla and galla, were in course of time
contracted into md and gd.
The latter fact is an illustration of the extent to which Accadian came
to be affected by phonetic decay, and thus a ready means provided for the
transition of an ideograph into the symbol of a syllabic sound. The more I
have investigated the phonology of the language the more I have been
astonished by the extraordinary extent to which the loss of sounds and
syllables was carried. Almost the only parallel I can find to it is in the
Mandarin dialect of China ; and the literary fossilisation of the language, and the
wide spread of education in a country where the very soil furnished materials
for supplementing the language of the ear by the language of the eye, had no
doubt much to do with this waste and wear of sounds. So also had the
contact of the language with the younger and more vigorous Semitic, which
tended to enrich itself at the expense of its older neighbour. Final sounds
were chiefly attacked, but initial and even medial letters were also dropped,
and it not unfrequently happened that two words of totally different origin
came to assume the same form, like our box or sound. Thus, as noticed
above, J*" was primarily a representation of "the tongue " (c=3), and as such
expressed the ideas of "voice," "calling," "assembly," and the like. When
inserted within the mouth (>-^p^) it denoted sometimes " the tongue,"
sometimes " speech," sometimes " a nation." Now both these characters
originally had the value eme, that being the Accadian word which expressed
' W. A. I. IV, 2, 5, 37, 3«.
154
LECTURE IX.
the ideas they stood for, but in course of time erne became attenuated to me.
On the other hand, there was another word mcs, (J*-*-*^) which signified
" many," and was expressed by a combination of the two characters J*— (now
pronounced mc) and <« [cs). Mes often denoted the plural ; mur, for instance,
being " a brick," w?/r-w^yy^>^ was called masuminnabi, or '' masu
twice;" t^ diliminnabi, or ''dili (>— ) twice." Where the form of a character
was changed by the addition of new wedges, the word gunu (" tailed "
probably), was used; e.g., ^^^yyt^ was Mggagunu, from "^yy^ ^^gg^ ; tW
was nindagunu, from ^ ninda. Sometimes gitu "drawn back," or "horizontal,"
was the epithet employed, as in K^^ i-gUu, and ^ s'a-gUn. The postposition
cu "to," occasionally makes its appearance, as in t^tlj gistarurassacu, ''gistar
' In his recent work Lcs Syllabaircs ciincifonncs {i^-]"]).
AFFINITIES OF ASSYRIAN. 1 55
plus urassa,'' or the 3rcl person singular of the aorist of the verb dii " to go,"
preceded by the pronoun ^ar *' what." Thus >-^^ is called gay-itu-cii-bat-idu
"what to itu (>-^^) bat (>-^) goes," i-du standing for in-dii. When three
characters are united together, the compound ideograph takes its name from
all of them ; thus ►f- >^}^^]r tH^yj is gadataccuru, as being a compound of
gada, tak, and urn. The primary object of thus naming the characters was
clearly to enable the scribe to write from dictation, and this fact will explain
several of the errors met with in the inscriptions. M. Lenormant has pointed
out, that in many cases it is only when we refer to the archaic Babylonian
form, that the compound character of the ideograph and the reason of its
name can be discovered, and since in some few cases even the archaic
Babylonian form does not afford an explanation of the name, it is evident
that the classification of the syllabary goes back to a very remote period
indeed. Many of the so-called syllabaries were drawn up merely for the sake
of classifying the characters and cataloguing their names, and they have to
be carefully distinguished from those other syllabaries in which the Accadian
word in the first column is interpreted by the Assyrian equivalent in the third
(or fourth). M. Lenormant has done good service in separating the two
kinds of syllabary, and printing them apart. In the case of the first kind, the
third column does not contain the Assyrian rendering of the Accadian word
in the first, but merely the Semitised form of the Accadian name of the
character. Thus dil = >^ = dilu, signifies that the Accadian name of >^
was dilj to which the Assyrians attached the vowel-ending of their nominative.
On the other hand, di-il ^ >— = na-bu means that the Accadian word dil,
represented by the character >^, signified "to proclaim," the Assyrian 7tabu.
It may be added that Mr. Smith found a fragment of one of those earlier
Babylonian syllabaries of which the Assyrian are but later copies ; if the
libraries of Babylonia are ever excavated we may expect to discover a
complete set.
Before concluding this Lecture, I would draw your attention to the illustra-
tion afforded by Assyrian of obscure words and ajra^ Xeyoixeva in the Old
Testament, as well as of words used by Rabbinical authors. I have already
alluded to the light thrown by the Assyrian estin (or estinu) "one," upon the
Hebrew ^ashte in "iJi^)/ "TIJ^I^, another instance would be the aVaf Xeyofxevov,
n^^i*, which occurs in Job ix. 26, and is exactly represented in Assyrian by
156 LECTURE IX.
abatu, " a ship." So, too, the Hebrew HnS and [^D, originally applied to
Assyrian officers, have received illustration from the Assyrian pakhat "a
governor," and 5^^;/// " a prefect." " Naturally Assyrian titles, like those of
the Rab-shakeh and the Tartan, have been cleared up. The first denoted the
Prime Minister or Grand Vizier, the second the Commander-in-chief of
Nineveh. Rab-shakeh is the Assyrian Rab-saki ** great one of the princes,"
the second part of the compound being of Accadian origin (5^/^ " head ") ;
Tartan is the Assyrian Turtannii, itself derived from the Accadian Tur-dan, or
" powerful prince." So, again, the meaning of the obscure word D^HX in
Isa. xiii. 21, translated "doleful creatures" in the Authorised Version, has
been determined by the corresponding Assyrian akhii, which represents the
Accadian lig-barra or " hyaena." As might have been expected, many words
of rare occurrence in Hebrew are met with plentifully in the inscriptions. I
need only refer to khuratsu, VT\T\, "gold," agammu {agdmit), DJi^, " a pool,"
or s'aniu I In (s'ainulu), hf2D, "an image."
The language of the Talmud and the Targums was of course largely
affected by that of Babylonia. Thus Dr. Delitzsch has pointed out * that the
Assyrian names of the four winds, iltanu or istann "the north," sutit "the
south," sadu " the east," and akharru " the west," are reproduced in the
Gemara under the forms of J^JTID^^, ^mt^, i^H^i^, and ^^m^^. Sadu " the
east" wind, originally signified the wind of "the mountains," the Accadian
sad, the mountains in question being those of Elam. Khazan, again, in the
sense of "governor," explains the ]Tn of the Mishna, which has now come to
mean " a leader " in prayer or singing in the Synagogue. Passing over words
like the Assyrian cissu (for cinsii), "multitude," which corresponds with the
Targumic Wl^ with a Ji^, or the doubling of "l, which was allowed both in
Assyrian and Babylonian Hebrew, we find the Assyrian katu " a hand," from
the Accadian kat, reappearing in the Talmudic ^np " a handle." Dr.
Neubauer's conjecture that the Talmudic "^"|"^3 " slave," is derived from "lO
"to sell," is confirmed by the Assyrian kinnatn "a female slave," which
probably goes back to the root Hip "to buy," and, as Harkavy has noticed,
■ Dr. Pusey has observed that the Hebrew "iDpp (Jer. li. 27; Nah. iii. 17), translated "captain" in the
Enghsh Version, is explained by the Accadian dhip-sar, " (man of) the written tablet(s)," or "scribe," which
was adopted by the Assyrians, and through them handed on to the Jews.
■ Assyrische Studicn, i. p. 140.
AFFINITIES OF ASSYRIAN. 1 57
the Talmudic ^^^^HJ " a gift," finds an explanation in the Assyrian nadanii
" to give," with *! instead of the usual ]n3, while the Targum uses 33J in the
sense of "uniting," like the Assyrian gabbu "all." A better etymology than
the Greek Xol/jlo? or At/xdy can be found for the Rabbinic odl in the Assyrian
lanias's'u [lamcidu) " a colossus," the origin of which is to be sought in the
Accadian lamma or lamdsi.
The recovery of the Assyrian language, in fact, is vindicating the Semitic
nationality of many Targumic and Talmudic words which it has been the
fashion to refer to a Greek source. Dr. Delitzsch observes ' that the
Aramaic N/^lSK is not a disguised form of the Greek efi^oXr), but has its
analogue in the Assyrian abiillu or " city-gate." So, too, i<^"'lP " sweet wine,"
is connected, not with the Greek KapoLvov but with the Assyrian caranu, and
OplD is to be referred rather to the Assyrian dhakadu^ "to arrange," than to the
Greek rafty.
But Assyrian lexicography is still in its infancy. We are still employed
in completing the grammar of the language, and here alone has anything like
success been obtained. An Assyrian grammar is possible, but not yet an
Assyrian dictionary. In the preceding course of Lectures I have endeavoured
to introduce you to the outlines and main features of that grammar, and to
smooth over the difficulties which beset the path of the beginner. How far I
may have succeeded is for you to say. In parting, I cannot refrain from
thanking you for the attention you have bestowed upon my efforts, and
expressing my gratification at the large and persevering group of students
that have accompanied me through the dry details of an extinct grammar.
Let us not forget that we are all learners together, and that the success which
has attended the present course of Lectures is the best possible augury for the
future progress and achievements of English Assyriology.
■ George Smith's Chalddische Genesis, p. 298.
THIS BOOK IS DUE ON THE LAST DATE
STAMPED BELOW
AN INITIAL FINE OF 25 CENTS
WILL BE ASSESSED FOR FAILURE TO RCTURN
THIS BOOK ON THE DATE DUE. THE PENALTY
WILL INCREASE TO 50 CENTS ON THE FOURTH
DAY AND TO $1.00 ON THE SEVENTH DAY
OVERDUE.
JUL 15 1944
*^<" IJ 10 4 4
" ? ]..o
JUN 3 1961
"'^^fs^^^eigTT^'
feS^ rrftM^'
LD 21-10m-5, '43 (6061s)
U.C.BERKELEY LIBRARIES
CD^b^a^l'=^3
'^9:m.-4eai/*amia' twa n